mead cohen berger shevtsova garfinkle michta grygiel blankenhorn bayles
Romney’s Magic Number: 3.12

[iframe src=””]

Via Meadia is not and never will be a site that follows every twist and turn of a presidential race. What candidate commits a forgettable gaffe, how a candidate’s team spins a disappointing debate or primary performance, what no-hoper GOP also-ran says mean things about another bit part player: that isn’t what we want to write about here.

This doesn’t mean we intend to rise sublimely above an election that engages the attention of the whole world; increasingly leaders and analysts all over the world will be trying to assess the chances of President Obama’s re-election as a factor in their planning. Financial markets will be watching as well; besides the presidential race, they will be looking to see what kind of Congress is likely to emerge from the fray and whether that Congress will be more able than the current assembly to put the country’s finances on a sustainable course.

Last week we introduced our election map, not as an attempt to forecast the (still wildly unpredictable and completely up for grabs) November result, but as a way to show graphically where the race is right now. Using the RCP “poll of polls” as a reference point, we compared the candidates’ poll standing now with the results of 2008, adjusted the electoral vote to reflect the results of the 2010 census, and showed the electoral map that those polls (with undecideds divided proportionately between the candidates) would produce in an election.

This week we’ve added a couple of new wrinkles. First, we have tweaked our approach to undecided voters, assuming that some of them will vote for third party or other candidates in November. For now, we are assuming no change in the third party vote between 2008 and 2012 and that the rest of the undecideds will break along the same lines as the rest of the public; we will test and review those assumptions as the election approaches and better data becomes available.

Second, we are adding the “magic number:” the national swing from the current leader (President Obama) to his opponent needed to change the election result. At the moment, that magic number is 3.12 percent: a swing of that size toward Romney would, distributed equally across the country, give the former Massachusetts governor a narrow 275-263 victory in the Electoral College.

The big difference: with that narrow swing in the popular vote, Florida, Virginia, Ohio and Colorado shift into the Romney column. (We aren’t calculating numbers for Rick Santorum at this point; at the moment Mitt Romney appears all but certain to be the Republican nominee.)  With a slightly larger swing, Romney’s electoral majority could widen substantially; Iowa, Minnesota, New Hampshire and Pennsylvania would all be in play.

Many readers took our first national map as a sign that President Obama was all but assured of a comfortable win. Perhaps — but a swing of a little over three percent is not hard to imagine.  It is also important to note that after more than three years in office, President Obama has lost ground since his last win. At the President’s current level of political support, North Carolina and Indiana, two normally red states that the then Senator Obama carried in 2008, have flipped back to the GOP.

The current polls if anything overstate the President’s strength. The latter stages of the GOP nomination battle have been tough, and the President enjoys the luxury of an opposition that is so busy with internecine strife that it cannot develop much less follow up on a consistent program of attacking the President’s record.

All things being equal, we would expect the race to tighten in the months to come. Without getting drawn into breathless blow-by-blow commentary we’ll continue to keep an eye on the big picture, and in the next couple of weeks will start to look at what current voter sentiments will do for the House and the Senate.

Features Icon
show comments
  • Francis Urquhart

    I think one great flaw in your calculation is spreading the undecided vote proportionally to the decided vote. They tend to swing to the challenger. As the election nears, overwhelming so.

    • Walter Russell Mead

      @Francis Urquhart: This far out, that is harder to assess. Since most voters aren’t really focused on the election yet, being undecided does not mean as much as it will later in the campaign. As we get closer, we will adjust the model to reflect changing probabilities — and of course, we will have better tools to assess the potential third party vote.

  • Kris

    … Furthermore, one of our interns has a head for numbers, so might as well use it.


  • John Burke

    Urquhart is right. The assumption that the incumbent will not win undecideds proportionally is rooted in the idea, born out through decades of survey experience, that voters have already decided pro or con about the incumbent; it’s the challenger they don’t know. This is especially true of Presidential elections since nearly everyone has an opinion about an incumbent President. This has nothing to do with how far off the election is, since the whole idea of a poll or average of current polls is that it’s a snapshot of what would happen if the election took place now.

    Moreover, you can’t assess the current state of play with only the matchup polls. A President’s approval numbers are likely to be more instructive about the outcome (if the election were held today). It is a good rule of thumb to say that an incumbent needs to be at 50 percent approval or very close — 48-49 — or is in trouble. Low 40s or below spells disaster. Mid 40s means the outcome is iffy. Obama is now at 47 in the RCP approval average.

    How should you apply the approval data? Basically, by putting a question mark over the outcome in states that might now be tilting a bit one way while there are good reasons to think it might tilt the other way.

    The problem with Mead’s map is that he seems determined to assign every state to one side or the other, despite the fact that the evidence available is not so definitive. I don’t know what he’s got against the concept of “tossup” but I think map’s like this one by Larry Sabato are more informative — and predictive:

    And what’s this 3.12 percent “magic number?” I guess it’s consistent with Mead’s assigning every electoral vote based on thinly sliced polling data. But does he honestly think he can state a percentage to the second decimal point that would translate to exactly 270 electoral votes?

  • Peter

    I will be quite surprised if there is the tremendous energy on Obama’s behalf that was present in the last election. He had an unbelievable turnout from minorities and the young. The only group where his support is about the same as in 2008 is the black community. On the other hand, there are a great many people who will be motivated to vote by their dislike of Obama. Conservatives may not be enthused about Romney, but they are dead set against Obama. They will be coming out this election.

  • WigWag

    “The big difference: with that narrow swing in the popular vote, Florida, Virginia, Ohio and Colorado shift into the Romney column” (Walter Russell Mead)

    Professor Mead’s map proves how daunting Romney’s task really is. If the rest of Mead’s map proves prescient, Obama could lose three of the four swing states the the Professor mentions and still win the election. If Obama wins Ohio and loses Colorado, Florida and Virginia he still beats Romney 281 electoral votes to 257 electoral votes.

    To make matters still worse for Romney, even if Obama loses two additional states that he should win, Iowa and New Hampshire (where Romney owns a large home), Obama still beats Romney 271 electoral votes to 267.

    This election is coming down to the results of three states; Florida, Ohio and Pennsylvania. Whichever candidate wins two of those three states is going to win. If there is a credible scenario where one of the candidates loses two of those three states and still wins the election, I can’t imagine what it is.

  • Chase

    Professor, you are underestimating how unpopular leverage buyout firms, like Bain, are in the Midwest, and among middle and working class voters everywhere. George W. Bush made his money in Baseball which, unlike big fiance, is a business with an all American image (even if the steroid scandals suggest that it doesn’t deserve that image).

    The secret to GOP success is to get a personable candidate that doesn’t look like s/he is backed by big money. Reagan and Bush 43 were absolute masters at this game. W might have been very rich, and his backers might have been super rich, but doggone it, he sure loved clearing brush on the weekends down in Texas. It will be much more difficult for Romney to pull this off. If anyone doubts me on this, google “Romney small varmints,” and enjoy the show.

  • Andrew Allison

    I suspect that a map of Romney vs Santorum would look very much like the one above. Some discussion of the implications, namely the down-ticked damage that a Romney candidacy would cause, would be interesting. The Republican party hopes that the anti-Obama feeling will cause voters to hold their noses and “do the right thing”. Thinking back to 1992, I’m not so sure.

© The American Interest LLC 2005-2017 About Us Masthead Submissions Advertise Customer Service