Fun With Party I.D.

'Fun With Party ID' takes political speculation to a new level with some innocent poll manipulation . . . all for the sake of looking at elections and polling results from a "what if?" perspective.

[partisanid-blog]

It's Public Policy Polling vs. The World With Regards to Gov. Bob McDonnell's Approval Ratings

Public Policy Polling stands out among polling firms for their partisan, often times combative nature.

In the Virginia Governor poll released by PPP yesterday, and referenced by this blog, there was an easily missed statistic regarding the job approval rating for Virginia’s Governor. According to the Democratic leaning pollster, Bob McDonnell (R) is above water with Virginia voters, but only somewhat:

Do you approve or disapprove of Gov. Bob McDonnell’s job performance? (May 24-26, 2014; 672 Virginia voters, 3.8% MoE) Approve  -  44% Disapprove  -  37% Not Sure  -  19%

This finding represents a net 6 point deterioration in approval for McDonnell’s job performance since PPP’s last poll in January (48/35%), and represents their lowest net rating for the Governor since July of 2012. But more significantly, the above finding represents a fairly significant departure from every other pollster to test the Virginia Governor’s job approval this year, or even since taking office in 2010. Just consider Quinnipiac University, which released a survey on Bob McDonnell’s job approval rating less than two weeks before PPP. They found McDonnell’s overall net job approval rating to be 3 TIMES higher than PPP (49/28% vs. 44/37%), while his rating among whites was DOUBLE PPP’s finding (50/29% vs. 48/36%). Meanwhile, there’s an astounding 30 point spread between Independents’ views on Gov. McDonnell’s job performance (54/27% per Quinnipiac, and 39/41% per PPP). But there’s no need to stop at Quinnipiac. Just 3 weeks ago, the Washington Post released a survey showing McDonnell with a net job approval rating 5 TIMES higher than PPP just found. And the same week, NBC/Marist found the same thing, as the chart below of McDonnell’s job approval rating since the start of this year notes:

McDonnell’s lowest rating this year besides PPP’s most recent survey came in January, when he hit 48/35%, from none other than….PPP. Every other polling firm this year has found McDonnell between +21 and +37 in terms of net job approval/disapproval. And the outlier-nature of PPP’s Virginia Governor job ratings doesn’t end with just this year:

The chart above documents every Virginia job approval survey taken by PPP since McDonnell took office in 2010. As you can see, he’s averaged a respectable 46/35% approval rating over 3.5 years in office. But that figure fails in comparison to McDonnell’s overall job approval average with all pollsters BESIDES PPP, which stands at a very enviable +28, or 57/29%:

So PPP hasn’t simply found McDonnell’s approval rating THREE TIMES lower than other pollsters this year. They’ve consistently found him with lower approval ratings for all 3.5 years of his Governorship…three times lower, in fact. And they’ve found it across various demographics as well. For example, McDonnell averages a +34 approval rating with white voters according to all other pollsters, whereas PPP finds him at just +20. And while he’s averaged a 58/28% rating with Independents among non-PPP pollsters, he’s at just 48/34% with PPP. What does all this mean? It means that you might want to exercise a bit of caution when considering PPP’s Virginia Governor job approval ratings. They’ve consistently found McDonnell significantly weaker than all other pollsters. Their findings might just be chalked up to random error, if they weren’t so consistent, and if they didn’t come from such an overtly partisan organization.

[partisanid-blog]

2009 VA Gov Electorate Voted 51-43% for McCain; PPP finds 2013 Electorate Giving 45% to Romney

John McCain, seen here with Va. Gov. Bob McDonnell in Oct. '09, wound up being a valuable asset to the campaign. Of the Virginians that voted on election day '09, 51% said they supported him in '08. Photo courtesy of Preston Gannaway & The Virginian-Pilot

Yesterday, Public Policy Polling released their second Virginia Governor poll of the year, with very little changed from the last survey in January. Terry McAuliffe (D) maintains his 5 point lead over Ken Cuccinelli (42-37%), while McAuliffe's net favorability rating fell slightly, and Cuccinelli's rose slightly. Oh, and in a rare move, the number of undecided voters actually increased from 13% to 21% since the last survey 5 months ago. At the same time, PPP finds the Virginia electorate remains largely unchanged as well. Five months ago, Democrats held a 3 point partisan identification advantage (35% Dem, 32% Rep, 32% Ind), which is now down to one point (34% Dem, 33% Rep, 33% Ind). The racial identification findings from January and now are virtually identical (74% white, 18% black, 8% other), as well as ideological I.D., age I.D., gender, and '2012 vote'. But how does the PPP electorate compare to the one found by exit polls in 2009, when Republican Bob McDonnell defeated Democrat Creigh Deeds in a landslide? In most ways, it's similar, with one big exception. White voters comprised a slightly larger slice of the electorate in '09 (78%) than what PPP finds now (74%), but not enough to significantly alter the poll results. The 2009 race featured an R+4 party I.D., while PPP finds D+1 (again, the difference is not enough to substantially change the survey result if reweighted). PPP also finds a larger gender turnout gap than seen in 2009, as well as a slightly older electorate, but nothing that would change the fundamentals of their topline. The one exception to the above paragraph can be seen in the crosstab regarding the 2012 vote. Often times, pollsters will ask respondents who they voted for in the previous presidential election, mostly to get an idea of how the pool of survey takers compares to the prior presidential electorate (this question was frequently used in survey analysis in the run-up to the South Carolina 1st Congressional District special election between Mark Sanford and Elizabeth Colbert-Busch to gain insight into the strength of GOP turnout). As has been noted here before, off-year, non-presidential races have a tendency (though not always) to attract less Democratic friendly electorates than in presidential years (see the 2010 vs. 2008 general election exit polls). But that is not the case in either of PPP's Virginia Governor polls. In both surveys, 49% of respondents say they voted for Barack Obama in 2012, while 45% claim they supported Mitt Romney. 6% say they supported someone else, or do not remember. That's an identical margin to the actual result in Virginia last November, when Obama carried the state 51-47%. The '2012 vote' finding may not seem so surprising until you consider what the 2009 Virginia Governor electorate looked like in comparison to the 2008 Presidential vote. That year, Virginia voters told exit pollsters they supported John McCain over Barack Obama in the 2008 presidential election by a fairly substantial 51-43% margin. Interesting thing is...John McCain LOST the state of Virginia in 2008 by an equally substantial 53-46%. In other words, the Virginia electorate shifted 15 points towards McCain voters between November 2008 and November 2009. Now obviously, non-presidential year electorates are not uniform. But the pattern of off-year elections in Virginia fearturing more Republican presidential voters than Presidential elections played out again in 2006 (unfortunately, there are no exit polls available for the 2005 and 2001 VA Governor races). That year, Democrat Jim Webb barely edged out George Allen for the Senate seat, at the same time as Virginia voters said they supported George W. Bush over John Kerry in the previous election by a 52-42% margin. The actual 2004 Virginia presidential result? Bush beat carry by 8 points. So 2006 counts as another non-presidential year contest where voters claimed to have voted for the Republican in the previous presidential election by a larger margin than they actually did. Understanding that no particular demographic crosstab should be considered in a vacuum, what would the PPP topline look like had they found an electorate that was equally more supportive of the Republican candidate from the previous presidential election as in 2009? In other words, how would the result change had PPP found the Virginia electorate voted 52% for Mitt Romney and 41% for Barack Obama (a net 15 point shift towards the Republican candidate, like seen in 2009):

Unlike the above mentioned minor differences between PPP's crosstabs and 2009 exit poll findings, reweighting the survey's '2012 vote' to mirror a 15 point Republican shift causes a drastic change in results. As opposed to McAuliffe being ahead 42-37%, the numbers are completely reversed, with Cooch winning 42%, to McAuliffe's 37%, a 10 point shift. PPP's '2012 vote' finding in the Virginia Governor's race is not set in stone. In fact, as we learned from the SC-1 special election, it can change drastically from week to week. But if their current finding holds, it would represent an electorate that is identical in make-up to the 2012 presidential election results, whereas the 2009 gubernatorial race saw an electorate that shifted 15 points in favor of the Republican candidate from the 2008 election. Whatever the true nature of who likely Virginia Governor voters supported in 2012 at this point, both the Democratic and Republican candidates are within 5 points of each other. With 5 months until election day, it's still anyone's race.

[partisanid-blog]

2009 VA Gov Electorate Voted 51-43% for McCain; PPP finds 2013 Electorate Giving 45% to Romney

John McCain, seen here with Va. Gov. Bob McDonnell in Oct. '09, wound up being a valuable asset to the campaign. Of the Virginians that voted on election day '09, 51% said they supported him in '08. Photo courtesy of Preston Gannaway & The Virginian-Pilot

Yesterday, Public Policy Polling released their second Virginia Governor poll of the year, with very little changed from the last survey in January. Terry McAuliffe (D) maintains his 5 point lead over Ken Cuccinelli (42-37%), while McAuliffe's net favorability rating fell slightly, and Cuccinelli's rose slightly. Oh, and in a rare move, the number of undecided voters actually increased from 13% to 21% since the last survey 5 months ago. At the same time, PPP finds the Virginia electorate remains largely unchanged as well. Five months ago, Democrats held a 3 point partisan identification advantage (35% Dem, 32% Rep, 32% Ind), which is now down to one point (34% Dem, 33% Rep, 33% Ind). The racial identification findings from January and now are virtually identical (74% white, 18% black, 8% other), as well as ideological I.D., age I.D., gender, and '2012 vote'. But how does the PPP electorate compare to the one found by exit polls in 2009, when Republican Bob McDonnell defeated Democrat Creigh Deeds in a landslide? In most ways, it's similar, with one big exception. White voters comprised a slightly larger slice of the electorate in '09 (78%) than what PPP finds now (74%), but not enough to significantly alter the poll results. The 2009 race featured an R+4 party I.D., while PPP finds D+1 (again, the difference is not enough to substantially change the survey result if reweighted). PPP also finds a larger gender turnout gap than seen in 2009, as well as a slightly older electorate, but nothing that would change the fundamentals of their topline. The one exception to the above paragraph can be seen in the crosstab regarding the 2012 vote. Often times, pollsters will ask respondents who they voted for in the previous presidential election, mostly to get an idea of how the pool of survey takers compares to the prior presidential electorate (this question was frequently used in survey analysis in the run-up to the South Carolina 1st Congressional District special election between Mark Sanford and Elizabeth Colbert-Busch to gain insight into the strength of GOP turnout). As has been noted here before, off-year, non-presidential races have a tendency (though not always) to attract less Democratic friendly electorates than in presidential years (see the 2010 vs. 2008 general election exit polls). But that is not the case in either of PPP's Virginia Governor polls. In both surveys, 49% of respondents say they voted for Barack Obama in 2012, while 45% claim they supported Mitt Romney. 6% say they supported someone else, or do not remember. That's an identical margin to the actual result in Virginia last November, when Obama carried the state 51-47%. The '2012 vote' finding may not seem so surprising until you consider what the 2009 Virginia Governor electorate looked like in comparison to the 2008 Presidential vote. That year, Virginia voters told exit pollsters they supported John McCain over Barack Obama in the 2008 presidential election by a fairly substantial 51-43% margin. Interesting thing is...John McCain LOST the state of Virginia in 2008 by an equally substantial 53-46%. In other words, the Virginia electorate shifted 15 points towards McCain voters between November 2008 and November 2009. Now obviously, non-presidential year electorates are not uniform. But the pattern of off-year elections in Virginia fearturing more Republican presidential voters than Presidential elections played out again in 2006 (unfortunately, there are no exit polls available for the 2005 and 2001 VA Governor races). That year, Democrat Jim Webb barely edged out George Allen for the Senate seat, at the same time as Virginia voters said they supported George W. Bush over John Kerry in the previous election by a 52-42% margin. The actual 2004 Virginia presidential result? Bush beat carry by 8 points. So 2006 counts as another non-presidential year contest where voters claimed to have voted for the Republican in the previous presidential election by a larger margin than they actually did. Understanding that no particular demographic crosstab should be considered in a vacuum, what would the PPP topline look like had they found an electorate that was equally more supportive of the Republican candidate from the previous presidential election as in 2009? In other words, how would the result change had PPP found the Virginia electorate voted 52% for Mitt Romney and 41% for Barack Obama (a net 15 point shift towards the Republican candidate, like seen in 2009):

Unlike the above mentioned minor differences between PPP's crosstabs and 2009 exit poll findings, reweighting the survey's '2012 vote' to mirror a 15 point Republican shift causes a drastic change in results. As opposed to McAuliffe being ahead 42-37%, the numbers are completely reversed, with Cooch winning 42%, to McAuliffe's 37%, a 10 point shift. PPP's '2012 vote' finding in the Virginia Governor's race is not set in stone. In fact, as we learned from the SC-1 special election, it can change drastically from week to week. But if their current finding holds, it would represent an electorate that is identical in make-up to the 2012 presidential election results, whereas the 2009 gubernatorial race saw an electorate that shifted 15 points in favor of the Republican candidate from the 2008 election. Whatever the true nature of who likely Virginia Governor voters supported in 2012 at this point, both the Democratic and Republican candidates are within 5 points of each other. With 5 months until election day, it's still anyone's race.

[partisanid-blog]

What Joe Miller's Return Means for the GOP: Alaska Senate could be the Christie v. Buono of 2014

Joe Miller and Lisa Murkowski debate just prior to the 2010 midterms. According to polling, Miller would lose to incumbent Mark Begich in a landslide in 2014. Photo by Al Grillo, courtesy of Corbis.

Caught up in a media world that that not only longs for, but demands competitive, theatrical political races, it's easy to gloss-over those contests lacking in the "drama" department. And what makes a better political contest than a close one? One where candidates are evenly matched, funded to the tee, and stir the emotions of their supporters and detractors? But alas, the soap-opera contest cannot always be. Landslides are occasionally inevitable (usually involving a particularly strong incumbent or an unusually weak or unpopular challenger). Sometimes, however, they're entirely avoidable. And yesterday, the 2014 Alaska Senate race took one big step closer to becoming one of those avoidable landslides for the GOP. Yes, the Alaska Senate race features a one-term, relatively unknown Senator defending his seat for the first time in a state with a decidedly Republican slant. It's certainly not a state where you'd expect a freshman Democrat, elected by one point in a wave-Presidential year, to be in a position to win a midterm reelection overwhelmingly. But fortunately for Senator Mark Begich (D-AK), the only Republican challenger making clear signals about running is a man who is despised by his state's general electorate, and underwater with members of his own party... Joe Miller. So what happened to the man who came painstakingly close to becoming Senator just two and a half years ago? What caused the guy who defeated incumbent Lisa Murkowski in a Republican primary to become so disliked within his own party? A series of ethical and ideological controversies that emerged near the tail-end of his almost-successful 2010 Senate campaign seem to be the catalyst, and his narrow Senate loss and subsequent law-suit to challenge the results did little to stop the bleeding. Though whatever the cause of the once-rising Conservative stars' fall, the consequences seem pretty clear. Alaskans say they would vote overwhelmingly to reelect Democrat Mark Begich if their second option is Republican Joe Miller. Not only that, but Alaska REPUBLICANS would just assume have anybody else - any potential Republican candidate - as their nominee in 2014. The chart below documents every potential Alaska 2014 Senate Republican primary match-up involving Joe Miller from every pollster to test the race thus far (and the pickings are pretty slim; they're limited to just Public Policy Polling and Harper Polling - both partisan firms). But the results are remarkably similar:

In nearly every possible Alaska GOP primary match-up one could imagine, no matter the ideology or number of candidates running against him, Miller winds up as the LAST choice of Alaska conservatives. Meanwhile Gov. Sean Parnell is the most formidable 2014 Alaska Republican in the primary, with Sarah Palin being the second most formidable. But it's not just the fact that Alaska Republicans would prefer other candidates to Joe Miller; it's the fact that they simply don't like the guy. The chart below documents the favorability rating of every possible GOP candidate taken this year, among Alaska Republicans only:

While potential GOP candidates Treadwell, Palin, and Parnell are well received by their party's voters, Miller struggles, never reaching much higher than 1/3 approval....again, among his OWN PARTY. But for the sake of argument, lets put aside for a moment the obvious difficulties Miller would face in a Republican Primary, and assume he makes it to the general to take on Sen. Begich. Then what? Well, things certainly don't get any rosier for Joe:

TWO pollsters have tested a Joe Miller vs. Mark Begich race, and BOTH find Begich winning in a cake-walk. And I'm not talking about your run of the mill double-digit landslide - I'm talking about a Chris Christie vs. Barbara Buono style 30 point rut. The good news for Republicans? As the chart above shows, it doesn't have to be this way. There are a number of possible Alaskans that could give Begich a run for his money. But simply put, Joe Miller is no longer one of them. There was a time when his future in Alaska politics looked bright. 2014 is NOT that time. 

[partisanid-blog]

What Joe Miller's Return Means for the GOP: Alaska Senate could be the Christie v. Buono of 2014

Joe Miller and Lisa Murkowski debate just prior to the 2010 midterms. According to polling, Miller would lose to incumbent Mark Begich in a landslide in 2014. Photo by Al Grillo, courtesy of Corbis.

Caught up in a media world that that not only longs for, but demands competitive, theatrical political races, it's easy to gloss-over those contests lacking in the "drama" department. And what makes a better political contest than a close one? One where candidates are evenly matched, funded to the tee, and stir the emotions of their supporters and detractors? But alas, the soap-opera contest cannot always be. Landslides are occasionally inevitable (usually involving a particularly strong incumbent or an unusually weak or unpopular challenger). Sometimes, however, they're entirely avoidable. And yesterday, the 2014 Alaska Senate race took one big step closer to becoming one of those avoidable landslides for the GOP. Yes, the Alaska Senate race features a one-term, relatively unknown Senator defending his seat for the first time in a state with a decidedly Republican slant. It's certainly not a state where you'd expect a freshman Democrat, elected by one point in a wave-Presidential year, to be in a position to win a midterm reelection overwhelmingly. But fortunately for Senator Mark Begich (D-AK), the only Republican challenger making clear signals about running is a man who is despised by his state's general electorate, and underwater with members of his own party... Joe Miller. So what happened to the man who came painstakingly close to becoming Senator just two and a half years ago? What caused the guy who defeated incumbent Lisa Murkowski in a Republican primary to become so disliked within his own party? A series of ethical and ideological controversies that emerged near the tail-end of his almost-successful 2010 Senate campaign seem to be the catalyst, and his narrow Senate loss and subsequent law-suit to challenge the results did little to stop the bleeding. Though whatever the cause of the once-rising Conservative stars' fall, the consequences seem pretty clear. Alaskans say they would vote overwhelmingly to reelect Democrat Mark Begich if their second option is Republican Joe Miller. Not only that, but Alaska REPUBLICANS would just assume have anybody else - any potential Republican candidate - as their nominee in 2014. The chart below documents every potential Alaska 2014 Senate Republican primary match-up involving Joe Miller from every pollster to test the race thus far (and the pickings are pretty slim; they're limited to just Public Policy Polling and Harper Polling - both partisan firms). But the results are remarkably similar:

In nearly every possible Alaska GOP primary match-up one could imagine, no matter the ideology or number of candidates running against him, Miller winds up as the LAST choice of Alaska conservatives. Meanwhile Gov. Sean Parnell is the most formidable 2014 Alaska Republican in the primary, with Sarah Palin being the second most formidable. But it's not just the fact that Alaska Republicans would prefer other candidates to Joe Miller; it's the fact that they simply don't like the guy. The chart below documents the favorability rating of every possible GOP candidate taken this year, among Alaska Republicans only:

While potential GOP candidates Treadwell, Palin, and Parnell are well received by their party's voters, Miller struggles, never reaching much higher than 1/3 approval....again, among his OWN PARTY. But for the sake of argument, lets put aside for a moment the obvious difficulties Miller would face in a Republican Primary, and assume he makes it to the general to take on Sen. Begich. Then what? Well, things certainly don't get any rosier for Joe:

TWO pollsters have tested a Joe Miller vs. Mark Begich race, and BOTH find Begich winning in a cake-walk. And I'm not talking about your run of the mill double-digit landslide - I'm talking about a Chris Christie vs. Barbara Buono style 30 point rut. The good news for Republicans? As the chart above shows, it doesn't have to be this way. There are a number of possible Alaskans that could give Begich a run for his money. But simply put, Joe Miller is no longer one of them. There was a time when his future in Alaska politics looked bright. 2014 is NOT that time. 

[partisanid-blog]

MA SENATE SPECIAL: Gabriel Gomez Is Doing Something Scott Brown Never Did in 2010 - He's Losing Momentum

Massachusetts has seen a fairly rapid turnover rate in the U.S. Senate after Sen. Ted Kennedy's death in 2009. Before then, the liberal lion served with Junior Senator John Kerry since 1985. Since then, Kerry served briefly as Senior Senator along with Scott Brown (R) from 2010-2013, when Kerry was nominated as Secretary of State, and Brown lost reelection. Now, polling indicates Elizabeth Warren and Ed Markey will become the new faces of the MA U.S. Senate delegation.

The Gabriel Gomez campaign for Massachusetts Senate was doused with cold water last week when both Emerson College and Public Policy Polling put a damper on what had been a slew of good polling news. While Gomez is still in a respectable position for a Republican running in one of the most liberal states in the Union, the polls are moving in the wrong direction from where he stood immediately after capturing his party's nomination for U.S. Senate last month. In the first four surveys released following his Republican primary victory on April 30, Republican candidate and ex-Navy SEAL Gabriel Gifford never trailed his Democratic challenger Edward Markey by more than 8 points, or by less than 3 points. That is, until a Suffolk University poll emerged about a week later showing the long-term Democratic Congressman with a fairly significant, and more typical lead over Gomez of 52-35%. As written about here at the time, the Suffolk survey appeared to be an outlier. Until then, the average of post-primary polling had Markey (D) up 45-40% on Gomez (R), while Gomez's numbers with women, men, Republicans, Democrats, Independents, whites, and non-whites were all within a strong to manageable range. But over the last two weeks, the Suffolk survey has grown to look less outlier-ish and more prophetic. Since then, Public Policy Polling has come out with their second post-primary Massachusetts Senate poll, and while the topline still looked 'okay' for Gomez, the trend lines looked bad. In just two weeks, Ed Markey (D) managed to nearly double his lead (from 44-40 to 48-41%). One week later, Emerson College was out with their 2nd post-primary survey, showing a different overall result from PPP, but with very similar trend lines. They too found Markey doubling his lead in just 3 weeks, from 42-36% to 45-33%. Suddenly, Suffolk University appeared to have company. So where did Gomez's support go? Where did he experience his largest drop offs? Emerson College and PPP agree in some places on the answer, while disagreeing in others. For example, both pollsters found the Republican Senate candidate's support among MEN slipping:

Emerson College found a net 14 point shift of support AWAY from Gomez among men, while PPP found a net 4 point shift away. Meanwhile the female vote remained largely static in both pollster's findings. In terms of the racial splits in voting, PPP and Emerson College diverge. Though both polls find Markey gaining in the overall topline result, his surge stems from an improved standing with non-white voters per PPP, and from white voters per Emerson College. In fact, PPP finds a massive shift in Markey's favor regarding the preference of non-white voters (Markey's level of support jumped from 47 to 55%, while Gomez's level of support collapsed from 34% to 15%). See the chart below:

Emerson College and PPP also present differing trends in terms of partisan support. The former shows Markey losing support among Democrats, while the latter shows him solidifying support. Emerson College finds Gomez's level of support among Republicans static, while PPP finds him solidifying GOP support. And while Emerson finds Gomez bleeding Independent support (going from +21 to +11 with this group), PPP finds him surging (jumping from +16 to +23):

Oh, and just for fun...what would the recent PPP and Emerson College polls had looked like had they both found partisan identification similar to the 2010 Brown v. Coakley numbers (35% Democrat, 12% Republican, 53% Independent):

*These pollsters asked respondents their party "registration," and not their party "identification."

Interestingly, if Emerson College and PPP came up with the same findings all over again, except partisan I.D. was identical to the 2010 Massachusetts Special Election, Gomez (R) would cut Markey's (D) lead in half. This is largely due to the fact that Emerson and PPP are finding less Independents in the 2013 electorate than exit polls measured in 2010. And given Gomez's substantial 56-33% lead with this group, obviously, he is going to benefit from their being more Independents in the electorate. In conclusion, there is almost no way to spin the poll results of the last two weeks in Gabriel Gomez's favor. After a seeming wave of momentum following his nomination, the race indeed appears to be settling in Ed Markey's favor. The only glimmer of hope for Gomez would be that PPP and Emerson are wrong about the partisan identification of the electorate, and old exit polling from 2010 is a more accurate representation. If that were the case, with 4 weeks until election day, Gomez could certainly make up what amounts to 3 and 6 point polling deficits. Also working in the Republicans favor is the fact at that this stage of the 2010 Special Senate election, loser Martha Coakley (D) still led Scott Brown in both Real Clear Politics and Pollster averages. In fact, no public poll showed Scott Brown with a lead over Martha Coakley until PPP's Jan 7-9, 2010 survey, just 11 days before the election. But unlike Gabriel Gomez, Scott Brown never lost his momentum in the race. Granted, he gained it late in the game. But once Brown gained momentum, he never ceded it, and he rode it to victory. The problem for Gomez may simply be that he peaked too soon.

[partisanid-blog]

MA SENATE SPECIAL: Gabriel Gomez Is Doing Something Scott Brown Never Did in 2010 - He's Losing Momentum

Massachusetts has seen a fairly rapid turnover rate in the U.S. Senate after Sen. Ted Kennedy's death in 2009. Before then, the liberal lion served with Junior Senator John Kerry since 1985. Since then, Kerry served briefly as Senior Senator along with Scott Brown (R) from 2010-2013, when Kerry was nominated as Secretary of State, and Brown lost reelection. Now, polling indicates Elizabeth Warren and Ed Markey will become the new faces of the MA U.S. Senate delegation.

The Gabriel Gomez campaign for Massachusetts Senate was doused with cold water last week when both Emerson College and Public Policy Polling put a damper on what had been a slew of good polling news. While Gomez is still in a respectable position for a Republican running in one of the most liberal states in the Union, the polls are moving in the wrong direction from where he stood immediately after capturing his party's nomination for U.S. Senate last month. In the first four surveys released following his Republican primary victory on April 30, Republican candidate and ex-Navy SEAL Gabriel Gifford never trailed his Democratic challenger Edward Markey by more than 8 points, or by less than 3 points. That is, until a Suffolk University poll emerged about a week later showing the long-term Democratic Congressman with a fairly significant, and more typical lead over Gomez of 52-35%. As written about here at the time, the Suffolk survey appeared to be an outlier. Until then, the average of post-primary polling had Markey (D) up 45-40% on Gomez (R), while Gomez's numbers with women, men, Republicans, Democrats, Independents, whites, and non-whites were all within a strong to manageable range. But over the last two weeks, the Suffolk survey has grown to look less outlier-ish and more prophetic. Since then, Public Policy Polling has come out with their second post-primary Massachusetts Senate poll, and while the topline still looked 'okay' for Gomez, the trend lines looked bad. In just two weeks, Ed Markey (D) managed to nearly double his lead (from 44-40 to 48-41%). One week later, Emerson College was out with their 2nd post-primary survey, showing a different overall result from PPP, but with very similar trend lines. They too found Markey doubling his lead in just 3 weeks, from 42-36% to 45-33%. Suddenly, Suffolk University appeared to have company. So where did Gomez's support go? Where did he experience his largest drop offs? Emerson College and PPP agree in some places on the answer, while disagreeing in others. For example, both pollsters found the Republican Senate candidate's support among MEN slipping:

Emerson College found a net 14 point shift of support AWAY from Gomez among men, while PPP found a net 4 point shift away. Meanwhile the female vote remained largely static in both pollster's findings. In terms of the racial splits in voting, PPP and Emerson College diverge. Though both polls find Markey gaining in the overall topline result, his surge stems from an improved standing with non-white voters per PPP, and from white voters per Emerson College. In fact, PPP finds a massive shift in Markey's favor regarding the preference of non-white voters (Markey's level of support jumped from 47 to 55%, while Gomez's level of support collapsed from 34% to 15%). See the chart below:

Emerson College and PPP also present differing trends in terms of partisan support. The former shows Markey losing support among Democrats, while the latter shows him solidifying support. Emerson College finds Gomez's level of support among Republicans static, while PPP finds him solidifying GOP support. And while Emerson finds Gomez bleeding Independent support (going from +21 to +11 with this group), PPP finds him surging (jumping from +16 to +23):

Oh, and just for fun...what would the recent PPP and Emerson College polls had looked like had they both found partisan identification similar to the 2010 Brown v. Coakley numbers (35% Democrat, 12% Republican, 53% Independent):

*These pollsters asked respondents their party "registration," and not their party "identification."

Interestingly, if Emerson College and PPP came up with the same findings all over again, except partisan I.D. was identical to the 2010 Massachusetts Special Election, Gomez (R) would cut Markey's (D) lead in half. This is largely due to the fact that Emerson and PPP are finding less Independents in the 2013 electorate than exit polls measured in 2010. And given Gomez's substantial 56-33% lead with this group, obviously, he is going to benefit from their being more Independents in the electorate. In conclusion, there is almost no way to spin the poll results of the last two weeks in Gabriel Gomez's favor. After a seeming wave of momentum following his nomination, the race indeed appears to be settling in Ed Markey's favor. The only glimmer of hope for Gomez would be that PPP and Emerson are wrong about the partisan identification of the electorate, and old exit polling from 2010 is a more accurate representation. If that were the case, with 4 weeks until election day, Gomez could certainly make up what amounts to 3 and 6 point polling deficits. Also working in the Republicans favor is the fact at that this stage of the 2010 Special Senate election, loser Martha Coakley (D) still led Scott Brown in both Real Clear Politics and Pollster averages. In fact, no public poll showed Scott Brown with a lead over Martha Coakley until PPP's Jan 7-9, 2010 survey, just 11 days before the election. But unlike Gabriel Gomez, Scott Brown never lost his momentum in the race. Granted, he gained it late in the game. But once Brown gained momentum, he never ceded it, and he rode it to victory. The problem for Gomez may simply be that he peaked too soon.

[partisanid-blog]

ABC / Washington Post House Ballot Poll Stands Out In More Ways Than One

Photo courtesy of http://historicalromneys.tumblr.com/

You know a poll finding might be a little screwy when the actual pollster releases the information with an explanation or caveat. And that's exactly what happened on Thursday morning when Gregory Holyk and Gary Langer of Langer research wrote this in explaining the substantial Democratic lead on the 2014 generic House ballot question in ABC/Washington Post's new national survey:

While it’s far too early to handicap the 2014 election in any serious way, registered voters currently favor the Democratic candidate over the Republican in their congressional district by 48-40 percent, the largest Democratic midterm advantage since 2006. The party saw steep losses in the 2010 midterms. The result almost entirely reflects a current Democratic advantage in partisan affiliation. Among registered voters in this survey, 33 percent identify themselves as Democrats, 22 percent as Republicans; most of the rest are independents, and they split evenly in their 2014 preference. Democrats outnumbered Republicans by 6 points in the 2012 presidential election and by 7 points in 2008, but they were even in the last two midterms – meaning the Democratic advantage holds only if their midterm turnout improves dramatically.

What Mr. Langer is saying here is not that their finding is incorrect (though that is certainly possible), but rather that if it is accurate, it would break precedent. And as I've pointed out on this blog before, he's right. A Democratic partisan identification advantage over Republicans of 11 points is just very, very unlikely in a midterm election. In fact, partisan identification in every midterm election since 1990 has ranged between D+3 and R+1, a fairly tight window of difference. Moreover, a glance at the 2014 House Ballot polling over the last month shows ABC/WaPo is a bit of an outlier, both in terms of party I.D., and the overall topline result:

Because Rasmussen Reports tracks the generic House ballot question once a week, their findings are listed as a monthly average for the sake of the chart.

As you can see, the average of pollsters shows the Democrats with just a 3 point advantage over Republicans on the generic House ballot, not 8 points, and with a 7 point party I.D. advantage, not 11 points. So not only is the Washington Post poll an outlier in the scheme of historical 2nd-term midterm election results, it's an outlier when compared to recent House ballot polling. But pointing that out apparently gets you labeled a poll truther or "unskewer" by the poll snobs at Huffington Post's Pollster. The NRCC noted all of the things I mentioned above on Thursday in a post meant to call into question ABC/WaPo's House ballot finding. In mocking fashion, Mark Blumenthal, without dealing at all with the merits of the NRCC's arguments with the respect to the poll, essentially dismissed their concerns as conspiratorial. It's one thing to arbitrarily change polling data and insist your results are correct (as famed unskewer Dean Chambers did). It's another to simply point out bizarre or unique polling data (as the NRCC was doing). In light of the poll-snobbery from Blumenthal, and why not...lets "unskew" that interesting ABC/WaPo House ballot finding to see what they would have found, all other findings remaining the same, if party I.D. was the same as the last Presidential race (38% Democrat, 32% Republican, and 29% Independent), or the same as the last midterm race (35% Democrat, 35% Republican, 29% Independent).

So there you have it. Shockingly (look, I can be snarky too!), had the poll NOT found a historical Democratic partisan ID advantage over the Republicans, we would have a much closer race on our hands. When the numbers are reweighted to the 2012 Presidential electorate finding (and remember, midterms electorates tend to favor Republicans more than Democrats as opposed to Presidential electorates), the overall Democratic lead drops to 5 points. When the ABC/WaPo poll is reweighted to the party ID from the last midterm election in 2010, the race is essentially tied, with the Democrats dropping to a one point overall advantage. On a final note, I couldn't help but notice ABC/WaPo's final survey prior to the 2010 midterm, released on October 28, 2010 (5 days before the election). They had the Democrats leading Republicans on the House ballot 49-44%. The actual result 5 days later? Republicans won a 51-45% popular vote victory, making ABC/WaPo off by a net 11 points. Making them HORRIBLY wrong. Oh, and the party I.D. on that poll? You guessed it. D+10. The party ID on actual election day? D+0.

[partisanid-blog]

ABC / Washington Post House Ballot Poll Stands Out In More Ways Than One

Photo courtesy of http://historicalromneys.tumblr.com/

You know a poll finding might be a little screwy when the actual pollster releases the information with an explanation or caveat. And that's exactly what happened on Thursday morning when Gregory Holyk and Gary Langer of Langer research wrote this in explaining the substantial Democratic lead on the 2014 generic House ballot question in ABC/Washington Post's new national survey:

While it’s far too early to handicap the 2014 election in any serious way, registered voters currently favor the Democratic candidate over the Republican in their congressional district by 48-40 percent, the largest Democratic midterm advantage since 2006. The party saw steep losses in the 2010 midterms. The result almost entirely reflects a current Democratic advantage in partisan affiliation. Among registered voters in this survey, 33 percent identify themselves as Democrats, 22 percent as Republicans; most of the rest are independents, and they split evenly in their 2014 preference. Democrats outnumbered Republicans by 6 points in the 2012 presidential election and by 7 points in 2008, but they were even in the last two midterms – meaning the Democratic advantage holds only if their midterm turnout improves dramatically.

What Mr. Langer is saying here is not that their finding is incorrect (though that is certainly possible), but rather that if it is accurate, it would break precedent. And as I've pointed out on this blog before, he's right. A Democratic partisan identification advantage over Republicans of 11 points is just very, very unlikely in a midterm election. In fact, partisan identification in every midterm election since 1990 has ranged between D+3 and R+1, a fairly tight window of difference. Moreover, a glance at the 2014 House Ballot polling over the last month shows ABC/WaPo is a bit of an outlier, both in terms of party I.D., and the overall topline result:

Because Rasmussen Reports tracks the generic House ballot question once a week, their findings are listed as a monthly average for the sake of the chart.

As you can see, the average of pollsters shows the Democrats with just a 3 point advantage over Republicans on the generic House ballot, not 8 points, and with a 7 point party I.D. advantage, not 11 points. So not only is the Washington Post poll an outlier in the scheme of historical 2nd-term midterm election results, it's an outlier when compared to recent House ballot polling. But pointing that out apparently gets you labeled a poll truther or "unskewer" by the poll snobs at Huffington Post's Pollster. The NRCC noted all of the things I mentioned above on Thursday in a post meant to call into question ABC/WaPo's House ballot finding. In mocking fashion, Mark Blumenthal, without dealing at all with the merits of the NRCC's arguments with the respect to the poll, essentially dismissed their concerns as conspiratorial. It's one thing to arbitrarily change polling data and insist your results are correct (as famed unskewer Dean Chambers did). It's another to simply point out bizarre or unique polling data (as the NRCC was doing). In light of the poll-snobbery from Blumenthal, and why not...lets "unskew" that interesting ABC/WaPo House ballot finding to see what they would have found, all other findings remaining the same, if party I.D. was the same as the last Presidential race (38% Democrat, 32% Republican, and 29% Independent), or the same as the last midterm race (35% Democrat, 35% Republican, 29% Independent).

So there you have it. Shockingly (look, I can be snarky too!), had the poll NOT found a historical Democratic partisan ID advantage over the Republicans, we would have a much closer race on our hands. When the numbers are reweighted to the 2012 Presidential electorate finding (and remember, midterms electorates tend to favor Republicans more than Democrats as opposed to Presidential electorates), the overall Democratic lead drops to 5 points. When the ABC/WaPo poll is reweighted to the party ID from the last midterm election in 2010, the race is essentially tied, with the Democrats dropping to a one point overall advantage. On a final note, I couldn't help but notice ABC/WaPo's final survey prior to the 2010 midterm, released on October 28, 2010 (5 days before the election). They had the Democrats leading Republicans on the House ballot 49-44%. The actual result 5 days later? Republicans won a 51-45% popular vote victory, making ABC/WaPo off by a net 11 points. Making them HORRIBLY wrong. Oh, and the party I.D. on that poll? You guessed it. D+10. The party ID on actual election day? D+0.

[partisanid-blog]

Despite rising job approval, Kay Hagan puts up her worst numbers yet against unsettled NC GOP field in new PPP poll

While some things have changed since 1998, a lot about the North Carolina electorate hasn't, according to Public Policy Polling crosstabs. That year, fmr. Sen. Lauch Faircloth (R-NC) (left) lost after one term in the U.S. Senate to newcomer John Edwards (D-NC). Photos courtesy of Ray Lustig / The Washington Post and Brooks Kraft/Corbis Sygma.

  For all the chatter about the Republican's current inability to recruit a quality candidate to challenge North Carolina's junior Senator in the 2014 election, a new Public Policy Polling survey offers the state GOP with a glimmer of hope. Out of the Democratic pollster's six polls on the race since December 2012, a Republican has finally managed to TIE Sen. Kay Hagan (D), despite Hagan reaching her highest job approval rating yet (46/40%). The chart below documents the results of PPP's 2014 NC Senate poll since December 2012, including only the candidates surveyed in the most recent poll:

Unfortunately for pollsters and researchers, exit polling in non-presidential year North Carolina Senate races is hard to come by (in 2010, there was no exit polling because Richard Burr led Elaine Marshall consistently throughout the campaign; in 2006, there was no NC Senate race; in 2002, major malfunctions with VNS led to exit polls being off, and as a result, incredibly difficult to come by. 1998 was the last NC Midterm with readily available exit poll data). This fact is important because without such data, it can be difficult for pollsters to weight the expected electorate. Nevertheless, a frequently stated rule of thumb regarding midterms is that they typically feature a more white, less ethnically diverse, older, and generally more Republican type of voter than seen in Presidential election years. So how do PPP's most recent findings regarding the potential 2014 electorate stack up as compared to the 2012 NC presidential electorate, and the 1998 Midterm electorate? At least in terms of racial identification, the PPP crosstabs reveal an electorate that looks very similar to the one seen in 1998, when Democratic pariah John Edwards, then just beginning his political career and still well-liked, defeated 1-term incumbent Republican Senator and pig-farmer Lauch Faircloth, 51-47%. That year, 76% of the electorate identified as white, 22% identified as black, and 2% Hispanic/other. Flash forward 16 years to PPP's May survey, and the electorate is 76% white, 18% black, and 6% Hispanic/other. In terms of partisan identification, age, and gender, the crosstabs diverge from the '98 Edwards/Faircloth race, at times resembling the 2012 electorate, and at times resembling nothing North Carolina electoral politics has experienced yet. The chart below displays the North Carolina electorate in every election featuring a readily available exit poll since 1992, broken down by gender identification, age identification, and partisan identification:

* 1992 CNN North Carolina exit polling divided age groups above 45 years of age differently than in later years.

Looking at gender identification, PPP finds North Carolina with the 2nd highest gender gap since 1992, with voters identifying as 56% female and 44% male (identical to turnout in the last presidential race). Such a gap is obviously beneficial to Sen. Hagan, as the Democrat leads with women in every potential match-up by as little as 8 points, and as much as 17 points. In NC's last midterm with exit poll data (1998), he gender gap was only 3 points, with 51% of voters being women, and 48% being men. Just for fun, the two far right column in the chart below documents how the PPP poll result would change in the event each match-up is reweighted to a more even split between men and women (along the lines of the 1998 midterm race):

As you can see, while the reweighting benefits Republicans slightly, the change is not significant. Shifting your eyes to the middle columns in the chart above, you'll see that while reweighting the PPP poll to 1998 gender identification benefited Kay Hagan's Republican opponents, reweighting to 1998 age identification does the opposite. That's because PPP found a decidedly older 2014 electorate than the one measured by VNS exit polling in 1998. And because Hagan performs better among younger voters, her vote margin against the Republicans would obviously grow when reweighting to 1998. Finally, the biggest difference between PPP's findings and 2012/1998 NC election exit polling was partisan identification. As the middle chart above indicates, PPP's D+12 finding represents the Democrats biggest advantage over the Republicans in any North Carolina election since 1992. In fact, it is actually more Democratic than either the 2012 Presidential election (D+6) OR the 1998 Edwards v. Faircloth election (D+7). It also goes a long way in explaining how Kay Hagan and Cherie Berry (R) can be TIED overall, while Berry performs better with her base, better with the opposing party, and leads among Independents 56-36%. So how would the PPP results change if they had found partisan identification identical to the 1998 midterm election? See the 4th column from the right in the chart below:

Had PPP found a party ID more along the lines of the 2012 Presidential election or the 1998 midterm, all other PPP findings remaining the same, Kay Hagan would have trailed her Republican opponents by as much as 3 points (47-44% against Cherie Berry), and would lead them by as much as 7 points (46-39% against Lynn Wheeler). As I note in most of these posts, the election is over a year away, and public opinion will certainly move in various directions between now and then. I'm certainly not suggesting that anything PPP is finding today is set in stone. Who knows how Benghazi, the IRS and AP scandals, or an improving economy will affect voter attitudes 18 months from now? But comparing different poll variables to past elections can provide us with a sort of snapshot of candidate floor/ceilings for a particular period of time in the overall race. As for now, Kay Hagan looks to be in pretty good shape for reelection. Her approval rating is fair and rising, and she's led all of her potential Republican rivals in every PPP poll since last December, with one exception. But she's still below 50% job approval, and seldom cracks the 50% threshold with any of her likely rivals in head-to-heads. And perhaps most importantly: she probably won't be competing in as friendly an electorate as the one she rode in on in 2008.

[partisanid-blog]

Despite rising job approval, Kay Hagan puts up her worst numbers yet against unsettled NC GOP field in new PPP poll

While some things have changed since 1998, a lot about the North Carolina electorate hasn't, according to Public Policy Polling crosstabs. That year, fmr. Sen. Lauch Faircloth (R-NC) (left) lost after one term in the U.S. Senate to newcomer John Edwards (D-NC). Photos courtesy of Ray Lustig / The Washington Post and Brooks Kraft/Corbis Sygma.

  For all the chatter about the Republican's current inability to recruit a quality candidate to challenge North Carolina's junior Senator in the 2014 election, a new Public Policy Polling survey offers the state GOP with a glimmer of hope. Out of the Democratic pollster's six polls on the race since December 2012, a Republican has finally managed to TIE Sen. Kay Hagan (D), despite Hagan reaching her highest job approval rating yet (46/40%). The chart below documents the results of PPP's 2014 NC Senate poll since December 2012, including only the candidates surveyed in the most recent poll:

Unfortunately for pollsters and researchers, exit polling in non-presidential year North Carolina Senate races is hard to come by (in 2010, there was no exit polling because Richard Burr led Elaine Marshall consistently throughout the campaign; in 2006, there was no NC Senate race; in 2002, major malfunctions with VNS led to exit polls being off, and as a result, incredibly difficult to come by. 1998 was the last NC Midterm with readily available exit poll data). This fact is important because without such data, it can be difficult for pollsters to weight the expected electorate. Nevertheless, a frequently stated rule of thumb regarding midterms is that they typically feature a more white, less ethnically diverse, older, and generally more Republican type of voter than seen in Presidential election years. So how do PPP's most recent findings regarding the potential 2014 electorate stack up as compared to the 2012 NC presidential electorate, and the 1998 Midterm electorate? At least in terms of racial identification, the PPP crosstabs reveal an electorate that looks very similar to the one seen in 1998, when Democratic pariah John Edwards, then just beginning his political career and still well-liked, defeated 1-term incumbent Republican Senator and pig-farmer Lauch Faircloth, 51-47%. That year, 76% of the electorate identified as white, 22% identified as black, and 2% Hispanic/other. Flash forward 16 years to PPP's May survey, and the electorate is 76% white, 18% black, and 6% Hispanic/other. In terms of partisan identification, age, and gender, the crosstabs diverge from the '98 Edwards/Faircloth race, at times resembling the 2012 electorate, and at times resembling nothing North Carolina electoral politics has experienced yet. The chart below displays the North Carolina electorate in every election featuring a readily available exit poll since 1992, broken down by gender identification, age identification, and partisan identification:

* 1992 CNN North Carolina exit polling divided age groups above 45 years of age differently than in later years.

Looking at gender identification, PPP finds North Carolina with the 2nd highest gender gap since 1992, with voters identifying as 56% female and 44% male (identical to turnout in the last presidential race). Such a gap is obviously beneficial to Sen. Hagan, as the Democrat leads with women in every potential match-up by as little as 8 points, and as much as 17 points. In NC's last midterm with exit poll data (1998), he gender gap was only 3 points, with 51% of voters being women, and 48% being men. Just for fun, the two far right column in the chart below documents how the PPP poll result would change in the event each match-up is reweighted to a more even split between men and women (along the lines of the 1998 midterm race):

As you can see, while the reweighting benefits Republicans slightly, the change is not significant. Shifting your eyes to the middle columns in the chart above, you'll see that while reweighting the PPP poll to 1998 gender identification benefited Kay Hagan's Republican opponents, reweighting to 1998 age identification does the opposite. That's because PPP found a decidedly older 2014 electorate than the one measured by VNS exit polling in 1998. And because Hagan performs better among younger voters, her vote margin against the Republicans would obviously grow when reweighting to 1998. Finally, the biggest difference between PPP's findings and 2012/1998 NC election exit polling was partisan identification. As the middle chart above indicates, PPP's D+12 finding represents the Democrats biggest advantage over the Republicans in any North Carolina election since 1992. In fact, it is actually more Democratic than either the 2012 Presidential election (D+6) OR the 1998 Edwards v. Faircloth election (D+7). It also goes a long way in explaining how Kay Hagan and Cherie Berry (R) can be TIED overall, while Berry performs better with her base, better with the opposing party, and leads among Independents 56-36%. So how would the PPP results change if they had found partisan identification identical to the 1998 midterm election? See the 4th column from the right in the chart below:

Had PPP found a party ID more along the lines of the 2012 Presidential election or the 1998 midterm, all other PPP findings remaining the same, Kay Hagan would have trailed her Republican opponents by as much as 3 points (47-44% against Cherie Berry), and would lead them by as much as 7 points (46-39% against Lynn Wheeler). As I note in most of these posts, the election is over a year away, and public opinion will certainly move in various directions between now and then. I'm certainly not suggesting that anything PPP is finding today is set in stone. Who knows how Benghazi, the IRS and AP scandals, or an improving economy will affect voter attitudes 18 months from now? But comparing different poll variables to past elections can provide us with a sort of snapshot of candidate floor/ceilings for a particular period of time in the overall race. As for now, Kay Hagan looks to be in pretty good shape for reelection. Her approval rating is fair and rising, and she's led all of her potential Republican rivals in every PPP poll since last December, with one exception. But she's still below 50% job approval, and seldom cracks the 50% threshold with any of her likely rivals in head-to-heads. And perhaps most importantly: she probably won't be competing in as friendly an electorate as the one she rode in on in 2008.

[partisanid-blog]

Tea Party Revival? Two New Polls Provide Conflicting Results in the Wake of the IRS Scandal

The Tea Party was the runner-up 'Person of the Year' for Time Magazine in 2010. Tea Party members, politicians, and activists are pictured above, from left: Sharron Angle, William Temple, Scott Brown, Sal Russo, Rick Santelli, Glenn Beck, Rand Paul, Jim Demint, Sarah Palin (center), Ron Paul, Dick Armey, Stephen Broden, Michele Bachmann, Christine O'Donnell, & Amy Kremer (far right). Illustration courtesy of Finlay Mackay

Though never known for their immense popularity with the public as a whole, the Tea Party has experienced a ratings-dive since their hey-day before the 2010 midterm election. Fortunately for them, there is some new poll data that indicates the recent IRS scandal may be providing the movement with a sympathy bump in terms of public perception. In 2010, pollsters found the anti-tax, anti-spending conservative group with an overall 33/34% favorable/unfavorable rating, while roughly the same portion of Americans either had no opinion, or refused to state their opinion. At the same time, 29% said they supported the movement, while 26% opposed it. Yet since those days when the stock market rose and fell in wild swings in a single day, when the unemployment rate peaked at a 10%, when wages plunged to a 20-year year low, public sentiment soured on the always alluring, yet always divisive grassroots movement. Following the Republican's 2010 midterm landslide, the Tea Party saw their evenly-divided favorability ratings drop to a decidedly negative 30/43% average. And though polling has been somewhat limited since the 2012 election, they've averaged an even less stellar 32/48% in the 6 months since November. Meanwhile, the margin of those saying they "support" rather than "oppose" the movement dropped 6 points since 2010. But in two surveys out after news broke on May 10th that the IRS targeted conservative groups, it appears the Tea Party is receiving a bit of a sympathy boost as a result of the perceived slight. CNN is the first to release poll findings showing the Tea Party with a notable gain in public perception. While a plurality of Americans still say they view the Tea Party unfavorably (45%), that number is down from 48% in CNN's poll two months earlier. The number of Americans viewing the Tea Party favorably has seen an even bigger jump, with 37% now viewing the group favorably, vs. 28% in March. Most of that improvement is seen among Democrats and Independents, not Republicans, oddly enough. ABC/Washington Post joined the fray yesterday with another survey showing the often demonized Tea Party with a respectable 40/43 favorability rating. But does polling truly indicate an emerging Tea Party resurgence?  It depends on which of the two post-IRS scandal pollsters you believe most: CNN or ABC/Washington Post. If you believe CNN, there is some definite movement in favor of the Tea Party as compared to two months ago. The chart below looks at every CNN survey on the Tea Party's favorability rating since 2011, immediately following their takeover of the House of Representatives:

The Tea Party's favorable/unfavorable rating has improved a net 12 points since March 2013, the best numbers measured by CNN since 2010. But while the new ABC/Washington Post poll numbers may sound like an improvement for the GOP, they're actually little-changed from the last survey measuring support/opposition for the Tea Party:

In fact, support for the Tea Party is slightly down from ABC/WaPo's last survey 8 months ago, in spite of the recent IRS scandal. Two caveats apply regarding the ABC/WaPo poll: 1) They have not surveyed support/opposition to the Tea Party for about 1 year until recently, which makes it hard to determine whether or not their recent 40/43% supportive finding is an increase or decrease from recent public sentiment. Though at least in terms of the numbers we have to work with, it appears Tea Party support is DOWN from their last poll before the 2012 election. 2) Question wording matters, and asking a poll respondent whether they view the Tea Party favorably or unfavorably, or whether they support or oppose them, can produce seemingly conflicting trends. Evidence of this can be seen in the two survey's topline results; poll respondents seem much more willing to rate the Tea Party unfavorable while also saying they support the group, as opposed to rating the group favorably, OR saying they oppose it - in other words, voters are more evenly split between the support/oppose question than they are the favorable/unfavorable question. In the end, the very limited post-IRS scandal polling we have suggests that Americans are viewing the Tea Party more favorably and less unfavorably than they did prior to May 10th, the day the IRS scandal broke. But at the same time, there is no positive movement in the direction of those that support the group rather than oppose it. It's often an easy excuse in poll analysis to say it will take more data to see whether or not disdain for the IRS has created a lasting improvement in perceptions of the Tea Party, but that is absolutely the case here. The Tea Party will need more than a single CNN data-point before anticipating a 2010-like renewal.

[partisanid-blog]

Tea Party Revival? Two New Polls Provide Conflicting Results in the Wake of the IRS Scandal

The Tea Party was the runner-up 'Person of the Year' for Time Magazine in 2010. Tea Party members, politicians, and activists are pictured above, from left: Sharron Angle, William Temple, Scott Brown, Sal Russo, Rick Santelli, Glenn Beck, Rand Paul, Jim Demint, Sarah Palin (center), Ron Paul, Dick Armey, Stephen Broden, Michele Bachmann, Christine O'Donnell, & Amy Kremer (far right). Illustration courtesy of Finlay Mackay

Though never known for their immense popularity with the public as a whole, the Tea Party has experienced a ratings-dive since their hey-day before the 2010 midterm election. Fortunately for them, there is some new poll data that indicates the recent IRS scandal may be providing the movement with a sympathy bump in terms of public perception. In 2010, pollsters found the anti-tax, anti-spending conservative group with an overall 33/34% favorable/unfavorable rating, while roughly the same portion of Americans either had no opinion, or refused to state their opinion. At the same time, 29% said they supported the movement, while 26% opposed it. Yet since those days when the stock market rose and fell in wild swings in a single day, when the unemployment rate peaked at a 10%, when wages plunged to a 20-year year low, public sentiment soured on the always alluring, yet always divisive grassroots movement. Following the Republican's 2010 midterm landslide, the Tea Party saw their evenly-divided favorability ratings drop to a decidedly negative 30/43% average. And though polling has been somewhat limited since the 2012 election, they've averaged an even less stellar 32/48% in the 6 months since November. Meanwhile, the margin of those saying they "support" rather than "oppose" the movement dropped 6 points since 2010. But in two surveys out after news broke on May 10th that the IRS targeted conservative groups, it appears the Tea Party is receiving a bit of a sympathy boost as a result of the perceived slight. CNN is the first to release poll findings showing the Tea Party with a notable gain in public perception. While a plurality of Americans still say they view the Tea Party unfavorably (45%), that number is down from 48% in CNN's poll two months earlier. The number of Americans viewing the Tea Party favorably has seen an even bigger jump, with 37% now viewing the group favorably, vs. 28% in March. Most of that improvement is seen among Democrats and Independents, not Republicans, oddly enough. ABC/Washington Post joined the fray yesterday with another survey showing the often demonized Tea Party with a respectable 40/43 favorability rating. But does polling truly indicate an emerging Tea Party resurgence?  It depends on which of the two post-IRS scandal pollsters you believe most: CNN or ABC/Washington Post. If you believe CNN, there is some definite movement in favor of the Tea Party as compared to two months ago. The chart below looks at every CNN survey on the Tea Party's favorability rating since 2011, immediately following their takeover of the House of Representatives:

The Tea Party's favorable/unfavorable rating has improved a net 12 points since March 2013, the best numbers measured by CNN since 2010. But while the new ABC/Washington Post poll numbers may sound like an improvement for the GOP, they're actually little-changed from the last survey measuring support/opposition for the Tea Party:

In fact, support for the Tea Party is slightly down from ABC/WaPo's last survey 8 months ago, in spite of the recent IRS scandal. Two caveats apply regarding the ABC/WaPo poll: 1) They have not surveyed support/opposition to the Tea Party for about 1 year until recently, which makes it hard to determine whether or not their recent 40/43% supportive finding is an increase or decrease from recent public sentiment. Though at least in terms of the numbers we have to work with, it appears Tea Party support is DOWN from their last poll before the 2012 election. 2) Question wording matters, and asking a poll respondent whether they view the Tea Party favorably or unfavorably, or whether they support or oppose them, can produce seemingly conflicting trends. Evidence of this can be seen in the two survey's topline results; poll respondents seem much more willing to rate the Tea Party unfavorable while also saying they support the group, as opposed to rating the group favorably, OR saying they oppose it - in other words, voters are more evenly split between the support/oppose question than they are the favorable/unfavorable question. In the end, the very limited post-IRS scandal polling we have suggests that Americans are viewing the Tea Party more favorably and less unfavorably than they did prior to May 10th, the day the IRS scandal broke. But at the same time, there is no positive movement in the direction of those that support the group rather than oppose it. It's often an easy excuse in poll analysis to say it will take more data to see whether or not disdain for the IRS has created a lasting improvement in perceptions of the Tea Party, but that is absolutely the case here. The Tea Party will need more than a single CNN data-point before anticipating a 2010-like renewal.

[partisanid-blog]

McAuliffe pulls away from Cuccinelli as Quinnipiac finds a 2012 style gubernatorial electorate

Terry McAuliffe and Creigh Deeds during a 2009 Democratic Primary debate for Governor. Photo courtesy of A.P.

Quinnipiac University is out with their fifth survey of the 2009 Virginia Governor's race, this time showing Democrat Terry McAuliffe with his highest survey margin yet over Republican Ken Cuccinelli.

If the election for Governor were being held today, and the candidates were Terry McAuliffe the Democrat and Ken Cuccinelli the Republican, for whom would you vote? Terry McAuliffe (D)  --  43% Ken Cuccinelli  (R)  --  38% Unsure/Other  --  20%

The results may surprise some, considering the barrage of negative press McAuliffe has received since Quinnipiac's last poll in March. But either Virginia voters are tuning out stories about McAuliffe's preference for Washington fundraisers over the birth of his children, or they just don't care. Because the Democrat has seen a 5 point gain in his support since March, while the Cooch has dropped 2 points. So what gave McAuliffe his largest lead of the Virginia Quinnipiac survey yet, especially in light of the perceived negative press? An initial glance at the crosstabs do little to explain the mystery, as both McAuliffe and Cuccinelli are winning similar portions of their own base, and Independents. In fact, Cuccinelli is actually doing slightly better than McAuliffe with his base, the opposing party, and Independents. But a look at the demographic and partisan make-up of Quinnipiac poll respondents sheds light on McAuliffe's rise. Simply, the current electorate is decidedly 2012-esque. The chart below documents every Quinnipiac Virginia Governor survey taken this season, including the partisan identification of the respondents of each poll.

As you can see, the most recent survey found Democrats with a 10 point partisan identification advantage over Republicans. That advantage is strong enough to give McAuliffe a 5 point lead over his challenger, despite both candidates performing equally well among their own party's base and Independents. Not only does this month's Quinnpiac poll feature the largest party I.D. advantage for Democrats to date, but if it holds through November, it would represent a net 14 point shift from the last Virginia Governor's race in 2009. That year, Republican Governor Bob McDonnell defeated his Democratic opponent Creigh Deeds in a near 20-point landslide, and Republicans turned out in much larger numbers than Quinnipiac now finds. In fact, that year, it was the Republicans that held a 4 point partisan I.D. advantage over the Democrats (37% R, 33% D, 30% I). What would the new Quinnipiac poll look like had their findings detected an electorate as Republican as 2009, all other findings remaining the same?

Not surprisingly, a 14 point shift in party I.D. (from D +10 to R +4), produces a net 10 point shift in Ken Cuccinelli's favor (from a 43-38% deficit, to a 45-40% lead). Of course, normal caveats apply. Pollsters find what they find (with the exception of a couple firms that do some of their own weighting - Rasmussen Reports and PPP, to be specific). But it's undeniable that if D +10 holds, it would be a remarkable turnaround from four years ago. A similar phenomenon can be seen with respect to Quinnipiac's most recent racial identification finding:

The chart above tracks the racial identification findings of all Quinnipiac Virginia Governor surveys, and finds Virginia Gubernatorial voters to be 70% white on average, and 30% non-white. Their most recent survey pegged racial identity at 71% white, 29% non-white. Like partisan identification, this finding represents a pretty large departure from the 2009 Governor's race. Than, 78% of voters identified as white, with just 22% identifying as non-white. Had Quinnipiac found a similar electorate in their last poll, McAuliffe would lead Cuccinelli by just 1 point, and not by a 43-38% margin:

It's impossible to know what kind of electorate we'll see in November 2013,  but the odds are that it will not be on par with Presidential turnout (as those elections tend to feature a younger, more ethnically diverse pool of voters). Yet Quinnipiac is currently finding just that - a gubernatorial electorate that looks awfully similar to 2012 (70/30% white/non-white; D+7 party I.D.). Such a scenario is possible, but Terry McAuliffe will need a very, very good night to pull those numbers off in a non-presidential, odd-year election.

[partisanid-blog]

McAuliffe pulls away from Cuccinelli as Quinnipiac finds a 2012 style gubernatorial electorate

Terry McAuliffe and Creigh Deeds during a 2009 Democratic Primary debate for Governor. Photo courtesy of A.P.

Quinnipiac University is out with their fifth survey of the 2009 Virginia Governor's race, this time showing Democrat Terry McAuliffe with his highest survey margin yet over Republican Ken Cuccinelli.

If the election for Governor were being held today, and the candidates were Terry McAuliffe the Democrat and Ken Cuccinelli the Republican, for whom would you vote? Terry McAuliffe (D)  --  43% Ken Cuccinelli  (R)  --  38% Unsure/Other  --  20%

The results may surprise some, considering the barrage of negative press McAuliffe has received since Quinnipiac's last poll in March. But either Virginia voters are tuning out stories about McAuliffe's preference for Washington fundraisers over the birth of his children, or they just don't care. Because the Democrat has seen a 5 point gain in his support since March, while the Cooch has dropped 2 points. So what gave McAuliffe his largest lead of the Virginia Quinnipiac survey yet, especially in light of the perceived negative press? An initial glance at the crosstabs do little to explain the mystery, as both McAuliffe and Cuccinelli are winning similar portions of their own base, and Independents. In fact, Cuccinelli is actually doing slightly better than McAuliffe with his base, the opposing party, and Independents. But a look at the demographic and partisan make-up of Quinnipiac poll respondents sheds light on McAuliffe's rise. Simply, the current electorate is decidedly 2012-esque. The chart below documents every Quinnipiac Virginia Governor survey taken this season, including the partisan identification of the respondents of each poll.

As you can see, the most recent survey found Democrats with a 10 point partisan identification advantage over Republicans. That advantage is strong enough to give McAuliffe a 5 point lead over his challenger, despite both candidates performing equally well among their own party's base and Independents. Not only does this month's Quinnpiac poll feature the largest party I.D. advantage for Democrats to date, but if it holds through November, it would represent a net 14 point shift from the last Virginia Governor's race in 2009. That year, Republican Governor Bob McDonnell defeated his Democratic opponent Creigh Deeds in a near 20-point landslide, and Republicans turned out in much larger numbers than Quinnipiac now finds. In fact, that year, it was the Republicans that held a 4 point partisan I.D. advantage over the Democrats (37% R, 33% D, 30% I). What would the new Quinnipiac poll look like had their findings detected an electorate as Republican as 2009, all other findings remaining the same?

Not surprisingly, a 14 point shift in party I.D. (from D +10 to R +4), produces a net 10 point shift in Ken Cuccinelli's favor (from a 43-38% deficit, to a 45-40% lead). Of course, normal caveats apply. Pollsters find what they find (with the exception of a couple firms that do some of their own weighting - Rasmussen Reports and PPP, to be specific). But it's undeniable that if D +10 holds, it would be a remarkable turnaround from four years ago. A similar phenomenon can be seen with respect to Quinnipiac's most recent racial identification finding:

The chart above tracks the racial identification findings of all Quinnipiac Virginia Governor surveys, and finds Virginia Gubernatorial voters to be 70% white on average, and 30% non-white. Their most recent survey pegged racial identity at 71% white, 29% non-white. Like partisan identification, this finding represents a pretty large departure from the 2009 Governor's race. Than, 78% of voters identified as white, with just 22% identifying as non-white. Had Quinnipiac found a similar electorate in their last poll, McAuliffe would lead Cuccinelli by just 1 point, and not by a 43-38% margin:

It's impossible to know what kind of electorate we'll see in November 2013,  but the odds are that it will not be on par with Presidential turnout (as those elections tend to feature a younger, more ethnically diverse pool of voters). Yet Quinnipiac is currently finding just that - a gubernatorial electorate that looks awfully similar to 2012 (70/30% white/non-white; D+7 party I.D.). Such a scenario is possible, but Terry McAuliffe will need a very, very good night to pull those numbers off in a non-presidential, odd-year election.

[partisanid-blog]

Is Alaska ready for a Palin comeback? They're not ruling it out, according to Harper Polling

Incumbent Senator Mark Begich (D) (left) could be facing fmr. Governor Sarah Palin (R) in his first reelection battle next November. Photo on left courtesy of Loren Holmes, photo on right courtesy of A.P.

GOP survey group Harper Polling is out with a new 2014 Senate Republican Primary Poll of Alaska, sponsored by the Tea Party Leadership Fund. And there's some encouraging news for Sarah Palin fans -- home-state Republicans still regard the former Governor and Vice Presidential nominee fondly, enough-so that they'd be willing to put her back in elected office next year:

If the Republican Primary election for U.S. Senator were held today, who would you vote for? Joe Miller, Sarah Palin, or Mead Treadwell? Sarah Palin  --  32% Mead Treadwell  --  30% Joe Miller  --  14% Not Sure  --  24%

In a one-on-one race with the only candidate to have announced an exploratory committee for 2014 so far, Joe Miller, Palin would lead her former protege 52-19%, with 29% undecided. Given the public opinion beating Palin took after resigning as Governor of Alaska in the Summer of 2009, followed by brutal media treatment in the aftermath of the Tuscon Arizona shooting of former Rep. Gabrielle Giffords (D), it may seem somewhat surprising to see Palin positioned to win her party's nomination for office again in her home-state. But among Alaska GOPers, her favorability rating is a strong 62/30%, she's viewed as the most likely to "fight for conservative values" by 20 points, and as the "strongest person to take on liberal Democrats in Washington" by 16 points.  But as the cross tabs will indicate, Harper Polling is finding a much more "conservative" GOP Senate primary electorate than the one PPP found in the immediate aftermath of 2010's contentious battle between Joe Miller and Lisa Murkowski. On August 24, 2010, Senator Lisa Murkowski (R) was NARROWLY defeated by Palin-endorsed Joe Miller for the GOP nomination for Senate, 51-49% (though Murkowski refused to concede the race until 1 week later). The rest was history, as Murkowski went on to wage a successful write-in campaign that November, defeating Miller (R) and McAdams (D) by a small margin. In the aftermath of that highly contentious Republican primary battle, PPP conducted something like an exit poll, though instead of polling Alaskan primary voters as they left the polling booths, they simply called them at their homes the following day. In that survey/exit poll of the 2010 Miller v. Murkowski primary battle, 59% of respondents identified themselves as Conservatives, 37% as Moderate, and 4% as Liberal. Now certainly, political environments can change, question wording can yield different results, and pollsters can just be wrong. And based on the disparate ideological I.D. findings of the two polling firms, one has to believe that one of those three things happened. While PPP found just 59% of the August 2010 Alaska GOP primary electorate to be conservative, 84% of Harper Polling respondents claimed to be conservative. While PPP measured 37% of the 2010 Alaska GOP primary electorate to be Moderate, Harper Polling found just 14% identifying as such. While PPP found 4% identifying as Liberal, Harper Polling found 1%. And as it's the habit of this blog, let's see what the new Harper Polling survey would have shown had they found a GOP primary electorate more like the one PPP says showed up when Miller narrowly defeated Murkowski in 2010:

Lt. Governor Sean Treadwell and former GOP Senate nominee Joe Miller both see their numbers improve under a 2010 turnout-scenario, while Sarah Palin's decline. Although Palin wins a plurality of Conservative Alaska GOP primary voters (35%), Treadwell doubles her margin among Moderate GOP voters. And since PPP found Conservatives making up a far smaller (and Moderates far larger) portion of the 2010 GOP primary electorate than Harper Polling, Treadwell moves from a 2 point deficit against Palin to a three point lead upon reweighting. Unfortunately, Harper Polling did not test Palin, Treadwell, or Miller against their Democratic opponent, Senator Mark Begich (elected in 2008). But they did  3 months ago. Then, Palin was able to make the race somewhat competitive (trailing the incumbent Begich 47-40%), while Miller would lose in a large landslide (52-29%). Lt. Gov. Treadwell was not tested against Begich in Harper's Jan. 29-30 survey. Now, exit polling in Alaska is hard to come by in presidential elections, much less primary or midterm exit polling. As a result, a comparison of Harper's January Alaska General Senate Election polling will have to be limited to 2008 and 2004 Presidential exit polls, as I was unable to locate any midterm Alaska exit polling. Either as a result of question working, pollster error, or the simple difference between midterm and Presidential electorates, Harper Polling finds a much more conservative and much less moderate 2014 Midterm Alaska electorate than exit pollsters found on Presidential election night 2008 and 2004 (much like they found regarding their primary electorate). That's why the chart below shows such drastic movement in favor of Sen. Mark Begich, in the event the Harper Poll were reweighted to 2008 or 2004 ideological I.D. findings:

Sarah Palin carries 63% of Conservative general electorate voters, but only manages 16% with Moderates, and 10% with Liberals. That may be enough for her to pull within single digits of Begich in a very Conservative electorate (like the one Harper Polling finds), but not if the electorate is more moderate (such as the electorates seen in 2008 and 2004). In fact, Palin does significantly worse against Begich when the Harper result is reweighted to past exit poll ideological findings, all other survey findings remaining the same (she trails by 24 and 26 points, instead of the originial 7 points). Yet interstingly enough, if the survey is reweighted to 2008 and 2004 partisan I.D. findings, Palin actually does slightly BETTER against Begich than the originial poll findings. That's because in both Presidential years, slightly more Republicans turned up to the polls than Harper Polling found in their survey, while a few less Independents and Democrats turned out. In conclusion, it would appear that while Sarah Palin's resignation did damage to her standing among Alaska voters (her overall favorability among ALL Alaskans is a poor 34/60%), not so much that they would be completely unwilling to put her back in elected-office. The numbers indicate that while she wouldn't be a shoe-in, she certainly has a strong chance. But as always the question with Sarah Palin...does she even want it? At least for now, the answer to that question looks like a no.

[partisanid-blog]

Is Alaska ready for a Palin comeback? They're not ruling it out, according to Harper Polling

Incumbent Senator Mark Begich (D) (left) could be facing fmr. Governor Sarah Palin (R) in his first reelection battle next November. Photo on left courtesy of Loren Holmes, photo on right courtesy of A.P.

GOP survey group Harper Polling is out with a new 2014 Senate Republican Primary Poll of Alaska, sponsored by the Tea Party Leadership Fund. And there's some encouraging news for Sarah Palin fans -- home-state Republicans still regard the former Governor and Vice Presidential nominee fondly, enough-so that they'd be willing to put her back in elected office next year:

If the Republican Primary election for U.S. Senator were held today, who would you vote for? Joe Miller, Sarah Palin, or Mead Treadwell? Sarah Palin  --  32% Mead Treadwell  --  30% Joe Miller  --  14% Not Sure  --  24%

In a one-on-one race with the only candidate to have announced an exploratory committee for 2014 so far, Joe Miller, Palin would lead her former protege 52-19%, with 29% undecided. Given the public opinion beating Palin took after resigning as Governor of Alaska in the Summer of 2009, followed by brutal media treatment in the aftermath of the Tuscon Arizona shooting of former Rep. Gabrielle Giffords (D), it may seem somewhat surprising to see Palin positioned to win her party's nomination for office again in her home-state. But among Alaska GOPers, her favorability rating is a strong 62/30%, she's viewed as the most likely to "fight for conservative values" by 20 points, and as the "strongest person to take on liberal Democrats in Washington" by 16 points.  But as the cross tabs will indicate, Harper Polling is finding a much more "conservative" GOP Senate primary electorate than the one PPP found in the immediate aftermath of 2010's contentious battle between Joe Miller and Lisa Murkowski. On August 24, 2010, Senator Lisa Murkowski (R) was NARROWLY defeated by Palin-endorsed Joe Miller for the GOP nomination for Senate, 51-49% (though Murkowski refused to concede the race until 1 week later). The rest was history, as Murkowski went on to wage a successful write-in campaign that November, defeating Miller (R) and McAdams (D) by a small margin. In the aftermath of that highly contentious Republican primary battle, PPP conducted something like an exit poll, though instead of polling Alaskan primary voters as they left the polling booths, they simply called them at their homes the following day. In that survey/exit poll of the 2010 Miller v. Murkowski primary battle, 59% of respondents identified themselves as Conservatives, 37% as Moderate, and 4% as Liberal. Now certainly, political environments can change, question wording can yield different results, and pollsters can just be wrong. And based on the disparate ideological I.D. findings of the two polling firms, one has to believe that one of those three things happened. While PPP found just 59% of the August 2010 Alaska GOP primary electorate to be conservative, 84% of Harper Polling respondents claimed to be conservative. While PPP measured 37% of the 2010 Alaska GOP primary electorate to be Moderate, Harper Polling found just 14% identifying as such. While PPP found 4% identifying as Liberal, Harper Polling found 1%. And as it's the habit of this blog, let's see what the new Harper Polling survey would have shown had they found a GOP primary electorate more like the one PPP says showed up when Miller narrowly defeated Murkowski in 2010:

Lt. Governor Sean Treadwell and former GOP Senate nominee Joe Miller both see their numbers improve under a 2010 turnout-scenario, while Sarah Palin's decline. Although Palin wins a plurality of Conservative Alaska GOP primary voters (35%), Treadwell doubles her margin among Moderate GOP voters. And since PPP found Conservatives making up a far smaller (and Moderates far larger) portion of the 2010 GOP primary electorate than Harper Polling, Treadwell moves from a 2 point deficit against Palin to a three point lead upon reweighting. Unfortunately, Harper Polling did not test Palin, Treadwell, or Miller against their Democratic opponent, Senator Mark Begich (elected in 2008). But they did  3 months ago. Then, Palin was able to make the race somewhat competitive (trailing the incumbent Begich 47-40%), while Miller would lose in a large landslide (52-29%). Lt. Gov. Treadwell was not tested against Begich in Harper's Jan. 29-30 survey. Now, exit polling in Alaska is hard to come by in presidential elections, much less primary or midterm exit polling. As a result, a comparison of Harper's January Alaska General Senate Election polling will have to be limited to 2008 and 2004 Presidential exit polls, as I was unable to locate any midterm Alaska exit polling. Either as a result of question working, pollster error, or the simple difference between midterm and Presidential electorates, Harper Polling finds a much more conservative and much less moderate 2014 Midterm Alaska electorate than exit pollsters found on Presidential election night 2008 and 2004 (much like they found regarding their primary electorate). That's why the chart below shows such drastic movement in favor of Sen. Mark Begich, in the event the Harper Poll were reweighted to 2008 or 2004 ideological I.D. findings:

Sarah Palin carries 63% of Conservative general electorate voters, but only manages 16% with Moderates, and 10% with Liberals. That may be enough for her to pull within single digits of Begich in a very Conservative electorate (like the one Harper Polling finds), but not if the electorate is more moderate (such as the electorates seen in 2008 and 2004). In fact, Palin does significantly worse against Begich when the Harper result is reweighted to past exit poll ideological findings, all other survey findings remaining the same (she trails by 24 and 26 points, instead of the originial 7 points). Yet interstingly enough, if the survey is reweighted to 2008 and 2004 partisan I.D. findings, Palin actually does slightly BETTER against Begich than the originial poll findings. That's because in both Presidential years, slightly more Republicans turned up to the polls than Harper Polling found in their survey, while a few less Independents and Democrats turned out. In conclusion, it would appear that while Sarah Palin's resignation did damage to her standing among Alaska voters (her overall favorability among ALL Alaskans is a poor 34/60%), not so much that they would be completely unwilling to put her back in elected-office. The numbers indicate that while she wouldn't be a shoe-in, she certainly has a strong chance. But as always the question with Sarah Palin...does she even want it? At least for now, the answer to that question looks like a no.