A survey of the most recent polls released over the last 72 hours shows Mitt Romney continuing to hold a lead over President Obama in the national race. However, President Obama continues to hold the advantage in the key state polls. There now appears a very real possibility that Romney will win the popular vote, and yet lose the election.
A Real Clear Politics average of ten polls from the last week has Romney gaining a bit in the national polls, with Mitt Romney garnering 47.9% support compared to 46.9% for President Obama. Last week Romney held just a 0.3 point lead on Obama in the RCP average. An average of the four tracking polls released over the last 72 hours shows Romney having the support of 48.75% of likely voters, compared to 46.75% supports for Obama.
An overview of the most recent national polls pitting Romney against Obama can be seen below. All polls are not created equally. The methods of each pollster vary, and many pollsters will adjust their numbers after their survey is complete. Ideally, these adjustments make the poll more accurate, but that is not always the case Finally, many pollsters also have a track record for either accurately predicting elections, or favoring one side over the other
As new polls are released, I will be providing continual updates, posting each poll along with some analysis on each pollster’s past reputation and performance. To receive more updates, follow me on Twitter or Facebook. I will also provide analysis on the latest trends.
Poll of 1500 “likely voters” taken from 10/24/2012 to 10/26/12
- Romney 50%
- Obama 46%
Analysis: Last week Rasmussen had Romney with a two point lead, so according to their data Romney has gained two points over that time. Rasmussen uses a fairly large sample and by “tracking” their results over multiple days the pollster should be able to produce more accurate results.
However, Rasmussen’s poll uses a “likely voter” sampling method, which tends to weigh Republicans slightly higher than Democrats. In the past, Rasmussen has been accused of producing biased results that tend to favor Republican candidates. Rasmussen does not include cell phone users, but attempts to compensate by using an “online survey tool to interview randomly selected participants from a demographically diverse panel.”
In 2008 Rasmussen’s final poll matched up well with the final result in the race, and its polls in the final month were fairly consistent. However, in 2010 a study done by the American Research Group found Rasmussen to be the least accurate of the eleven pollsters they evaluated. Rasmussen uses a sample that has “baseline targets” of 35.8% Republicans, 33.0% Democrats, and 31.2% “unaffiliated” voters, despite other surveys that show Democrats’ actually outnumbering Republicans in the general population.
Poll of 2700 likely voters taken from 10/19/2012 to 10/25/2012
- Romney 51%
- Obama 46%
Analysis: Last week Gallup had Romney with a seven point lead, and now has that lead down to five points. Gallup’s national poll has become by far the most controversial over the last week, as they are the only pollster showing Romney with a lead of more than two points. Some have claimed Gallup’s numbers stray too far from the average to really be taken seriously.
Gallup has, by far, the largest sample size, with 2700 likely voters, which decreases the margin of error for their polls to just +/-2.0%. Gallup also includes data from six days, which theoretically will even out their results, but the large date range could also be including old/bad data. Early in the election cycle Gallup sampled all registered voters, but now only include likely voters in their sample, which may aid Romney.
Gallup’s polls varied greatly in the 2008 race, at one point having McCain ahead by ten points before producing a final prediction with Obama up by eight points. Gallup has not released the internal breakdown of their sampling. Gallup was rated above average by Nate Silver in 2010.
Poll of 957 likely voters taken from 10/20/2012 to 10/25/2012
- Obama 47%
- Romney 45%
Analysis: The IBD/CSM/TIPP trackign poll is the only poll from the last week showing Obama with a lead nationally. One week ago IBD/CSM/TIPP’s tracking poll had Obama with a six point lead. Like Gallup, IBD/CSM/TIPP’s poll has drawn some skepticism for how far it varies from the mean.
In 2008, IBD/TIPP performed well in projecting the final result of the race, predicting a 7.2-point margin of victory for President Obama, which was the exact margin of victory for Obama once all the votes were countered. However, IBD/TIPP polls varied greatly in terms of consistency in the months leading up to the election, showing Obama with only a one-point lead as late as October 22, 2008. The two firms were not scored by Nate Silver in his 2010 June ratings.
NBC News/Wall Street Journal
Poll of 816 likely voters taken from 10/17/2012 to 10/20/2012
- Obama 47%
- Romney 47%
Analysis: NBC News/Wall Street Journal’s poll is the middle ground between the two extremes of Gallup and IBD/CSM/TIPP. They currently show the race tied. Their last poll in late September gave Obama a three point edge on Romney.
In 2008 NBC New/Wall Street Journal scored fairly well in terms of accuracy and consistency in predicting the actual results in the final month of the election. Nate Silver of The New York Times gave these pollsters an above average rating in 2010.
ABC News/Washington Post
Poll of 857 registered voters taken from 10/22/2012 to 10/25/2012
- Romney 49%
- Obama 48%
Analysis: The ABC News/Washington Post represents the middle ground in between the extremes of Gallup and IBD/CSM/TIPP over the last week and corresponds most closely to the average of polls. Eight days ago the same poll had Obama with a three point lead, so this pollster has Obama losing four points over that time.
In 2008 ABC News/Washington post performed poorly in terms of accuracy in predicting the presidential election results, but performed well in terms of consistency. The pollster is rated well above average by Nate Silver in his June 2010 ratings.
Poll of 839 registered voters taken from 10/19/2012 to 10/23/2012
- Romney 47%
- Obama 45%
Analysis: The Associated Press/AFK poll shows Romney with a two point lead in the race. The last survey this pollster released showed Obama with a one point lead in mid-September.
In his June 2010 ratings Nate Silver placed AP/GFK well above average, and ahead of both Rasmussen and Gallup. In 2008 the AP largely relied on polls from Gallup, so there is no data to analyze how the AP/GFK performed as a separate organization during that time.
Poll of 1,000 likely voters taken from 10/15/2012 to 10/18/2012
- Romney 49%
- Obama 47%
Analysis: POLITICO/GWU’s poll had Obama with a one point lead last week, so according to them the race has tilted toward Romney. This pollster did not release the party identification breakdown of their sample, but the likely voter model should favor Romney over Obama. This poll is the oldest of all the polls listed here, not inlcuding any daya from the last four days.
GWU Battleground scored well in terms of accuracy in predicting the final result of the 2008 election, but very poorly in terms of consistency in the month leading up the election.