Can Romney Win?

September 24, 2012

- See all 763 of my articles

We’re about six weeks away from election day, and the debates are just around the corner.  While the national polls show the race fairly tight, these polls are completely irrelevant.  A candidate gets the same number of electoral votes if he wins a state by one vote or by ten million.  Large margins may help the national polling numbers, but they don’t actually change the chance that the candidate will win.

For years. my go-to site has been  The site operator is very clear about the fact that he’s liberal, but I like the way the data is presented.  The site correctly predicted 48 of the 50 states in 2008, missing only Indiana and Missouri.

At this point, shows Obama with 328 electoral votes, Romney with 206, and New Hampshire tied.  Romney is expected to lose Massachusetts (where he was governor) convincingly.  That’s not surprising, considering how liberal the state is.  He’s also polling well behind in Michigan, where his father was governor.  While I haven’t researched this in great detail, I’d have to think that winning a presidential election while losing two states where you have deep personal ties would be a historic achievement.  Generally, a candidate can count on his home state to cast their votes for him.

One of my favorite features of the site is the tipping point state chart.  Essentially, this ranks the state from the ones where Obama is doing best (DC and Hawaii) to the states where he is doing worst (Utah and Wyoming).  If you’re a Romney fan, you can read from the bottom up.  The chart lets you see how far down (or up) a candidate must go in order to win.  In order for Romney to win, he’d have to pick up the tied state of New Hampshire, as well as Iowa, Nevada, Colorado, Florida, and Ohio – all states where Obama leads.  While it’s true that some of Obama’s leads are within the margin of error, this doesn’t mean that the poll results are wrong – it just means there is a greater likelihood that they are wrong.  Unless there is a systemic error affecting polling in multiple states, a candidate leading by less than the margin or error in large number of states is very likely to win a majority of those states.

How can Romney win?

  1. Explain Paul Ryan’s voucher system to senior citizens.  If seniors think that the Republicans are trying to take Medicare away, they could switch their votes or simply not vote at all.  Traditionally, this is a group that leans heavily Republican; it’s hard to imaging a Romney win that doesn’t include winning a majority of the senior vote.
  2. Capitalize on Obama gaffes in the debates.  However, it’s important to strike a chord with independent voters, and not simply preach to the choir.  This minds finding cases where the broader population disagrees with an Obama sentiment and hammering away.
  3. Getting out the vote.  A great way to overcoming a polling deficit is to simply get more of your party’s voters to the poll. 

Other news:


Tomorrow is the deadline for Todd Akin to petition a judge to remove his name from the ballot in the Missouri senate race.  If he does this (and the judge agrees) he would have to pay for ballots to be reprinted (possibly hundreds of thousands of dollars), but the Republicans would be able to put a different candidate on the ballot.  Claire McCaskill is comfortable head (although it’s not a blowout), and an Akin pullout would basically start with a clean slate.  I do wonder is McCaskill has been holding back a bit, in order to make Akin think he has a chance (rather than having him drop out and facing a stronger candidate).  Will McCaskill release the hounds on the 26th?

It appears that the Senate will add another independent to its ranks.  Former governor Angus King has a big lead over the Republican candidate.  Since King is expected to caucus with Democrats (although he has not publicly stated this), the Democrats are not supporting their party’s candidate, Cynthia Dill, for fear of splitting the vote between Dill and King and allowing the Republican to win.  Republican PACS have actually run ads in favor of Dill – trying to achieve just such a split.

Enhanced by Zemanta

4 Comments (+add yours?)

  1. Squeaky
    Sep 24, 2012 @ 12:38:37

    It’s going to be a HUGE undertaking if Romney is going to win and he can’t afford any more screw ups. Obama could fall flat on his face and still come out looking like a star gymnast while Romney stumbling seems to kill him.

    I agree that polls aren’t a great predictor, but you have to consider them to some degree. I’m disheartened about the polling, especially in my home state of Colorado. I was thinking about endeavors for the 2016 vote that I could start up and possibly run. Don’t put it past me to start a PAC for the next go round!

    I don’t want to be the one to admit defeat over 40 days from election, but the chance of Romney winning is very slim. It’s 4th quarter, 2:00 left on the clock and we’re behind 35-28. We’re at our own 20 with two timeouts left. The team is beat up and sore. It will surely take all the strength and determination to complete this favorably.

    The question is: What kind of QB is Romney and what kind of defense is Obama running? We’ll know in two minutes. (November 7th)


  2. kosmo
    Sep 24, 2012 @ 13:31:41

    The football analogy makes me think of the SNL spoof of the Eastwood ad.

    I don’t think polls as a whole are irrelevant; just the national ones.

    For example, let’s sample 100,000 voters in Colorado and 100,000 voters in Alabama. Obama would get 51,500 voters to 48,500 for Romney in Colorado. Romney would win Alabama 60,000 – 40,000. Overall, Romney gets 108,500 votes and Obama gets 91,500. That’s a 54.25% – 45.75% combined advantage in the two states.

    But at the end of the day, each would win 9 electoral votes. A casual observer might look at a 50%-50% tie in the national polls and assume the election is a coin flip – but in actuality it might be in the bag for one of the candidates.


  3. Martin Kelly
    Sep 25, 2012 @ 07:48:21

    One other thing we have to remember, as reported on NPR last week. The polsters are maintaining the 60-40 ratio of Dem-Rep from 2008 in their sampling when the last election (2010) showed a 48-48-4 Dem-Rep-Ind ratio. I have always maintained that the only important poll is the one on election day.

    Kos, you are absolutely correct in your views on the national poll. State by state is the only way to go, and even then, the biases of the polsters in both directions both in the question asked and the sample selections have made the daily polls less significant and more reported on.


  4. kosmo
    Sep 26, 2012 @ 08:59:38

    Hmm. I doubt that all the pollsters are using 2008 data for their sampling. That would be a very poor approach. They should be using active registration information to normalize their data.


Leave a Reply