Skip to main content

Something's not right with the Gallup tracking poll

Many Republicans are elated of the news that the Gallup tracking poll today has Mitt Romney ahead of Barack Obama by 7 points - 52 to 45. It's over, right? I'm sorry to disappoint some of my Republican friends, but I wouldn't get overly excited about that poll result just yet.

It is true that as of today, according to Gallup, Romney leads Obama by 7 points among likely voters - 52% to 45%. However, even in light of this, stats nerd Nate Silver gives the president a 65.7% chance of winning the election, compared to just a 34.3% chance for Romney. He bases these numbers on all the polls, trends, weighting for biases, history, along with other numbers with regard to the economy. Including this Gallup poll, he still gives President Obama close to a two-thirds chance of winning the election in close to three weeks. So, Gallup has Romney in a commanding lead, yet a stats expert only gives the Republican nominee about a one-thirds chance of winning on November 6th. What gives?

Something unusual is going on at Gallup. Let's go through some of their numbers here:

Among Likely Voters: Romney 52% Obama 45%

Approval Rating: Obama 50% approve 44% disapprove

Likely Female Voters in Swing States: Obama 49% Romney 48%

Among Female Swing State Voters - "What do you consider the most important issue in this election?"

1. Abortion - 39%
2. Jobs - 19%
3. Healthcare - 18%
4. The economy - 16%
5. Equal rights/pay/opportunity - 15%

Let's look at just those numbers for a moment. First off, according to this poll, Obama has an approval rating of 50% - +6 in approval rating. However, at the very same time, he is -7 to Romney, only garnering 45% of the vote. So, according to this poll, 5% of Romney voters approve of the president, yet won't be voting for him.

When it comes to women voters in swing states, Obama is just +1 on Romney, yet look at what these same women's most important issues are on voting day. Three of the top five they listed are strong points for the president when courting the woman's vote: Abortion, healthcare and equal rights/pay/opportunity, which comes out to be a whopping 72% of women in the survey. On the jobs and economy front, it's difficult to say which candidate would poll better among these women, but it's highly unlikely Romney would have an extraordinary lead in either category.

So, according to this poll, the president is rather popular and if just going by his approval ratings, he'd have an excellent chance of winning the election, yet when matched up with Mitt Romney, he loses 5 points. Also, while the president fares strongly with these swing state woman voters on the issues that are most important to them, he only leads Romney by 1 among them. Doesn't make a lot of sense, does it?

The Gallup poll makes even less sense when taking all other polls into account. Ipsos/Reuters has Obama up 3% overall. The Washington Post has him up 3 as well. RAND has the president up 6. TIPP has the two candidates in a tie. Right-leaning Rasmussen has Romney up by 2 (weighting their bias and that would come out to be about a +1 to +2 for Obama). The two candidates are tied in the PPP poll. Even if we don't weigh for bias, Obama has an average lead in those six polls by 1.7%. With Gallup, that drops to a 0.4% lead for the president. If we remove the two potential outliers (Gallup and RAND), Obama's lead would be at 0.8%. So, the fact that it appears Obama has a slight lead nationally, it makes the latest Gallup poll appear to be even more ridiculous.

Also, with regard to women voters in swing states, arguably the three biggest swing states in this year's election are Ohio, Florida and Virginia. In a series of recent polls released by NBC/Marist, Obama lead among likely women voters in the three before-mentioned states by the following:

Florida: Obama 54% Romney 43% (Obama +13%)

Virginia: Obama 54% Romney 42% (Obama +12%)

Ohio: Obama 54% Romney 42% (Obama +12%)

Given the Gallup poll which showed which issues were most important to women, these numbers would make far greater sense. Other reputable polls have shown similar numbers. So either the women in all other battleground states are overwhelmingly going to vote for Romney to place him at -1% overall among that demographic, or as I said at the outset - something's not right with the Gallup tracking poll. I'm going to say that the latter is much more likely than the former.

http://thehill.com/blogs/ballot-box/polls/262809-gallup-poll-romney-opens-up-7-point-lead-among-likely-voters

http://fivethirtyeight.blogs.nytimes.com/

http://www.gallup.com/home.aspx

http://www.usatoday.com/story/news/politics/2012/10/15/swing-states-poll-women-voters-romney-obama/1634791/

http://www.gallup.com/poll/158069/women-swing-states-gender-specific-priorities.aspx

http://maristpoll.marist.edu/1011-obama-and-romney-still-competitive-in-florida/

http://maristpoll.marist.edu/1011-romney-by-one-point-in-tight-contest-in-virginia/

http://maristpoll.marist.edu/1011-obama-still-leads-romney-in-ohio/

Comments

Popular posts from this blog

Boycotting jukeboxes because of TouchTunes

I love music and enjoy hitting the bar(s) over the weekend, so naturally, when the mood strikes me, I've never been coy about playing some songs on the jukebox. This past Thursday, a friend of mine turned 50, so several friends of her's, including myself, all met up to celebrate the occasion. At around 9:30, a friend of mine and I both chipped in $5 to play some songs on the jukebox. Four hours and 231 skips later, we gave up on hearing the songs we had selected, and went home knowing we had just wasted $5. This wasn't the first time such a thing had happened to me (and many others), and due to that, I'll be boycotting jukeboxes. Why? The scam known as TouchTunes. You see, here's how the plot typically breaks down. A person (or group of people) downloads the TouchTunes app on his/her phone, consumes one too many adult beverages, and due to this, has less care for spending extra money to hear the songs of their choosing right NOW. That's the thing with TouchTun

The difference between "looking" and "checking out"

I may be way off with these numbers, but it's my approximation that at least 75% of individuals whom are involved in a serious relationship feel it's perfectly acceptable to "check out" members of the opposite sex they're not involved with. Meanwhile, approximately 25% either don't feel this is acceptable or aren't sure about the matter. I hadn't thought about this matter for a while, but since I've been dating a woman for about 8 months, the topic has been pondered about some. When reading or hearing others discuss this very issue, I often times hear comments similar to the following: "It's human nature to look." "There's nothing wrong with checking others out. I'm sure he/she does it too!" "It's fine to do it. Just don't tell your boyfriend/girlfriend about it or do it in front of them!" "It's natural to find people attractive." When observing the array of comments, I i

The verdict is in. To no one's surprise, Jonathan Hoenig has been found guilty of being an idiot.

Just recently, when discussing the Michael Brown shooting and whether or not race had anything to do with it, Fox News contributor Jonathan Hoenig said, "You know who talks about race? Racists." One moment while I provide Mr. Hoenig with the well deserved slow-clap. :: slow-claps for two seconds :: So, that was quite the line by Mr. Hoenig, wasn't it? "You know who talks about race? Racists." Well, wasn't he just talking about race? So, by his own words, I guess that makes him a racist. Also, if he wants to be consistent, does this mean that people whom talk about gender are sexists and people whom talk about sexual orientation are homophobes? With that line of thinking, Hoenig would engage in the following back-and-forths: Hoenig: "So, who are you voting for?" A woman: "The Democratic candidate, because he's been adamant about his support for equal rights for women." Hoenig: "You sexist feminist nazi!"