Initially, many liberals panicked at the sight of the Gallup presidential election tracking poll and many conservatives jumped with glee. Republican presidential candidate Mitt Romney has held a steady lead in the Gallup poll and has been shown to lead President Barack Obama by as many as 7 points (52-45) within the past week or so. He's led the president by between roughly 3 and 7 points for quite some time. Of course the media has loved this poll for story, controversy, argument and ratings purposes - especially when Romney's lead was 7. However, as some began to wonder and then notice, it appeared as if, along with Rasmussen, Gallup's poll results didn't quite fit in with all the others, which have had the two candidates essentially tied. Numbers geeks such as Nate Silver began dismissing the Gallup poll as an outlier. This may have prompted Gallup to release more detailed information regarding their sample and they recently did just that.
Pretty much all the numbers regarding demographics are similar to the 2004 and 2008 elections. Whether it be with regard to gender, race, age, location or education, the numbers this year are almost identical to the previous two elections. However, there is one glaring difference in this year's Gallup sample compared to the two previous elections. Independent voters were sampled rather equally this year compared to the 2008 election (up slightly from 2004, but Independents have increased in quantity since then, so this is understandable). That hasn't been the case regarding Republican and Democratic samples, though. In 2008, Gallup's voter sample included 39% that were Democrat and 29% that were Republican. When including leaners, that brought the totals to 54% Democrat and 42% Republican. For this year's election, the sample includes 35% Democrat and 36% Republican, with Republicans holding a 49-46% advantage when leaners are included. That's a difference of -4 for Democrats, +7 for Republicans, -8 for Democrats/Democratic leaners and +7 for Republicans/Republican leaners for an overall difference of -11 for Democrats/+11 for Republicans and -15 for Democrats and Democratic leaners/+15 for Republicans and Republican leaners.
I don't care how enthused Republican voters are - that is an enormous switch in the sample. This is especially true since of the 25 polls which have been released this year regarding party affiliation, 24 of them showed Democrats with the advantage. In the most recent ten such polls, Democrats have held the advantage in all of them, by the average count of 39.0% to 33.6% - a +5.4% advantage for Democrats. There also doesn't appear to be a great enthusiasm gap amongst voters of the two parties. In a recent ABC/Washington Post poll, it showed that 62% of Romney's supporters were extremely enthused to vote for him and 62% of Obama's supporters were just as enthused to vote for him. With the enthusiasm gap being seemingly nil and Democrats holding a 5+% edge in registered/likely voters, it befuddles me why the Gallup tracking poll is giving Republicans a 3% edge here (8% difference from what registered/likely voter polls have indicated). Unfortunately for Gallup, Nate Silver, several talking heads and a decent majority of Democrats and liberals now see the poll as a joke and won't take it seriously until it lines up with the rest of the national tracking polls. Also unfortunately for Gallup, if it doesn't line up with the other tracking polls on the eve of the election and is the most inaccurate of them all (which is quite possible), they may be receiving some angry phone calls, e-mails, letters and telegraphs (okay, probably not this...) from Republicans. With Gallup temporarily suspending its poll in light of Hurricane Sandy, it'll be interesting to see if when they begin conducting and releasing polls again, if they stubbornly insist on being an outlier or alter their sampling any to fall in line with the rest of the national tracking polls.
http://www.gallup.com/poll/158399/2012-electorate-looks-like-2008.aspx
http://www.huffingtonpost.com/2009/06/23/party-id-rl_n_725932.html
http://thehill.com/blogs/ballot-box/polls/263663-romney-jumps-to-narrow-lead-in-wapo-abc-poll
Pretty much all the numbers regarding demographics are similar to the 2004 and 2008 elections. Whether it be with regard to gender, race, age, location or education, the numbers this year are almost identical to the previous two elections. However, there is one glaring difference in this year's Gallup sample compared to the two previous elections. Independent voters were sampled rather equally this year compared to the 2008 election (up slightly from 2004, but Independents have increased in quantity since then, so this is understandable). That hasn't been the case regarding Republican and Democratic samples, though. In 2008, Gallup's voter sample included 39% that were Democrat and 29% that were Republican. When including leaners, that brought the totals to 54% Democrat and 42% Republican. For this year's election, the sample includes 35% Democrat and 36% Republican, with Republicans holding a 49-46% advantage when leaners are included. That's a difference of -4 for Democrats, +7 for Republicans, -8 for Democrats/Democratic leaners and +7 for Republicans/Republican leaners for an overall difference of -11 for Democrats/+11 for Republicans and -15 for Democrats and Democratic leaners/+15 for Republicans and Republican leaners.
I don't care how enthused Republican voters are - that is an enormous switch in the sample. This is especially true since of the 25 polls which have been released this year regarding party affiliation, 24 of them showed Democrats with the advantage. In the most recent ten such polls, Democrats have held the advantage in all of them, by the average count of 39.0% to 33.6% - a +5.4% advantage for Democrats. There also doesn't appear to be a great enthusiasm gap amongst voters of the two parties. In a recent ABC/Washington Post poll, it showed that 62% of Romney's supporters were extremely enthused to vote for him and 62% of Obama's supporters were just as enthused to vote for him. With the enthusiasm gap being seemingly nil and Democrats holding a 5+% edge in registered/likely voters, it befuddles me why the Gallup tracking poll is giving Republicans a 3% edge here (8% difference from what registered/likely voter polls have indicated). Unfortunately for Gallup, Nate Silver, several talking heads and a decent majority of Democrats and liberals now see the poll as a joke and won't take it seriously until it lines up with the rest of the national tracking polls. Also unfortunately for Gallup, if it doesn't line up with the other tracking polls on the eve of the election and is the most inaccurate of them all (which is quite possible), they may be receiving some angry phone calls, e-mails, letters and telegraphs (okay, probably not this...) from Republicans. With Gallup temporarily suspending its poll in light of Hurricane Sandy, it'll be interesting to see if when they begin conducting and releasing polls again, if they stubbornly insist on being an outlier or alter their sampling any to fall in line with the rest of the national tracking polls.
http://www.gallup.com/poll/158399/2012-electorate-looks-like-2008.aspx
http://www.huffingtonpost.com/2009/06/23/party-id-rl_n_725932.html
http://thehill.com/blogs/ballot-box/polls/263663-romney-jumps-to-narrow-lead-in-wapo-abc-poll
Comments
Post a Comment