mead cohen berger shevtsova garfinkle michta grygiel blankenhorn
Where Things Stand: If the Presidential Election Were Held Today

[iframe src=’’ width=’100%’ height=’430px’]

So there it is: Via Meadia‘s best projection for the 2012 election results based on current polls.  You can compare Romney’s projected electoral vote with Santorum’s, and with President Obama’s actual results against Senator McCain in 2008. We’ve derived this from some quick and dirty back of the envelope guesses complex mathematical formulae using our state of the art computer facilities, including assigning undecideds to the two candidates (the poll numbers in the visual show the original RCP data without adjustments). The goal is less to predict the unpredictable results of an election that will be held almost eight months from now than to illustrate some of the dynamics that may help shape the race.

Even as a kid I was fascinated by the way the British polled their national elections. There are many differences between the two political systems, but one way to describe a general election in the UK is to say it is a contest to become prime minister in which seats in the House of Commons work like electoral votes in the US. It doesn’t matter how many popular votes you get; the party with the most seats in the House wins the election, and its leader becomes prime minister.

That poses a challenge for pollsters. With more than 600 seats in the House of Commons (the exact number has varied over time), it’s impossible to poll every constituency every day, but in the UK as in the US more than half the fun of democracy is watching elections, and so the Brits have developed a way to use national polls to predict electoral results in the House.

The theory is pretty simple; suppose in the last election the Labour Party got 42 percent of the vote nationally and that translated into 360 seats.  In some constituencies the Labour vote was higher than 42 percent and in some it was lower — and in some its margin of victory over other parties might have been five or six percent, in others one percent or less, and in still others it lost — again, coming closer in some constituencies than other.

Now suppose the current polls show Labour at 45 percent nationally — a three percent swing. By looking at the results from the last election you can make a quick and dirty estimate of how many seats Labour might win this time around. Just count up how many seats Labour lost by less than three percent last time around, and add those to its old total: you now can project what a three percent swing would do to the number of Labour-c0ntroled seats in the next House of Commons — and of course you can do the same thing for the other parties.

Looking at how changes in voting percentages change seat totals in Parliament is one of the favorite indoor sports for British political junkies; here is a BBC site where you can adjust national voting preferences and see how those changes would impact the House of Commons.

Via Meadia has been looking for a way to take note of the 2012 election without getting too deeply into the weeds, and for now at least we are going to use this quick and dirty British technique to show readers what current national opinion polls mean for the electoral college vote. In essence, the technique is to take a reasonably reliable poll result (we go with RCP’s “poll of polls”) and see how, once the undecideds are assigned, President Obama is doing compared to his performance in 2008. We assume that the national swing up or down in his support is distributed equally across all fifty states; our graphic has its limits, but over time it should give readers a good sense of where the race is going.

For now, we have tested President Obama against his two most likely GOP opponents: Governor Mitt Romney and Senator Rick Santorum. Using this method, it appears that if the election had been held today, President Obama would have handily won.  His percentage of the popular vote looks to be lower than it was in 2008, but this doesn’t seem to be impacting him as much in the electoral college as much as the raw polls might suggest.

The President is running slightly better against Santorum than he did against Senator McCain; indeed our poll projection shows the President carrying every state against Santorum that he did against McCain — and picking up Missouri as well. (Note that electoral vote totals reflect the changes mandated by the 2010 census.)

Governor Romney fares somewhat better: he holds Missouri and retakes North Carolina and Indiana.

Again, these maps are hardly predictions; most Americans won’t really start to make up their minds until much later in the year, and everything from the unemployment rate to the foreign situation will be factoring into those choices in ways that we cannot anticipate now.

Nevertheless, a few conclusions pop up from the charts:

  • the President is reasonably well positioned for re-election, but things could still go either way
  • the selection of Marco Rubio as a vice presidential candidate could shake up the race, making the GOP more competitive in New Mexico and Colorado while perhaps swinging Florida into the GOP column
  • GOP weakness in Ohio following Governor Kasich’s ill-judged attack on public labor unions could be a major factor in the fall
  • Virginia on the other hand could be less Democratic than it looks if worries about defense and intelligence cuts drive some of the northern Virginia vote back towards the GOP
  • While the drop off from his 2008 electoral support hasn’t yet hurt President Obama much in the electoral college, further erosion could change the electoral map dramatically. The President can’t afford much more slippage and hope to hold onto his job.

Let us know what you think of this new feature; if and when the polls change in any significant way, we will be back to show how those changes might be affecting the electoral map. We will also be looking at how swings in the national vote could affect the House and Senate makeup.

Features Icon
show comments
  • C. Phillips

    “Let us know what you think of this new feature”:

    It isn’t visible in my browser. Probably it won’t work without
    JavaScript. JavaScript is a serious security and privacy hazard
    in browsers, which I therefore keep off. (As just one example,
    keeping JavaScript off disables the tracking code associated
    with Facebook “like” buttons. It isn’t well known, but these
    buttons track you even if you never click on them and even if
    you don’t have a Facebook account.)

    Therefore, please change the coding to make it work
    without JavaScript. Thanks.

  • Damir Marusic

    Thanks for your feedback C. Phillips,

    You have to pick your poisons in this game. In this case, our options are JavaScript and Flash. Flash looks to be a dying technology at this point, especially when it comes to mobile devices. We’re a small shop here at the AI, and we can’t possibly support everything, so we’ve opted to go the JavaScript route. To set your mind at ease, this is the same JavaScript library that powers Google Maps.

    Privacy is nothing to scoff at, and the Internet is increasingly becoming a place where unscrupulous corporations just do what they please with your information. But JavaScript itself is not the problem—it’s an open technology standard which can be abused, but is not itself bad. Flash, on the other hand, (which also uses a version of JavaScript internally, by the way) is a closed system wholly owned by Adobe, with no shortage of security issues of its own.

    So I’d encourage you to give JavaScript another try. Have a look at which browsers offer you the best security settings and be careful about which cookies you accept from which sites. And surf with your eyes open.

  • David

    Is Obama really up 25 on Romney in Massachusetts? Seems like that projection and the Scott Brown v. Elizabeth Warren numbers couldn’t both be correct.

  • John Burke

    Honestly, I think this map is deeply flawed as a way to represent the current state of the race. Today’s RCP poll average has Obama ahead of Romney in Florida, yes — but by less than one point! Given everything else we know about the state of play, you would have to project Romney as the Florida winner or, at least, modify your map to include a tossup category.

    Similarly, Obama is today ahead of Romney in the RCP averages for Ohio and Virginia — but only by four points in both cases, well within the margins of error of the averaged surveys and with Obama getting only 45-46 percent of the straws, universally viewed as terrible for an incumbent. Given these factors and everything else we know — eg, Virginia’s voting history right up to and including 2009 and 2010 — you would have to put both these states in the tossup category. Ditto for New Hampshire and Colorado.

    All in all, your map is not very useful.

    • Walter Russell Mead

      @John Burke: If you mean the map isn’t useful at predicting what will happen in November, you are correct. But if the election were held today, a slim lead translates into a win and while there are always surprises, the candidate who polls show in the lead often does end up carrying a given state. But if you read the article carefully, you will see I think that it explicitly says this isn’t predicting the future and also points out that Obama’s slippage in the polls, if it continues, would result in a very different electoral map.

  • John Burke

    Addendum — Iowa, Minnesota, Nevada and Pennsylvania are tight, too. Wisconsin, Michigan, Oregon are leaning strongly to Obama now but Romney can still be competitive there. I don’t think such maps are very useful if they reflect merely the most recent intrastate polls.

  • WigWag

    My prediction is that Obama will lose Florida to Romney (he would beat Santorum in Florida) regardless of whether Romney selects Rubio. If Obama wins Florida I will be astounded.

    It is absolutely remarkable how the Republican Party is poised to snatch defeat from the jaws of victory in the presidential race.

  • John Burke

    @WRM — On the contrary, there is nothing in the history of modern American politics and polling that suggests that a slim polling lead translates into a win for the guy with the slim lead, if the election were held today. In Virginia, for example, Obama currently leads in the RCP average by four points, 46/42.Based on what we know from decades of polling, the bulk of the missing 12 percent will vote for the challenger, not the incumbent, because the latter is a far better known quantity and voters have already made up their minds about him. That’s why an incumbent at 46 is in such a terrible spot. He should be above 48 to have a shot (if the election were held today. Note how that principle would apply to the national numbers in the Obama-Romney matchup.

    But there is more. The 46-42 split is within the margins of error of all the polls in the average. This means that Romney could easily be ahead of Obama by a point or two in which case he would be a sure winner in today’s election.

    Finally — and this was my main point — in guessing how an election held today would turn out, current polling is extremely important but by no means the only input. For one thing, the six polls in the RCP average were taken between Jan. 19 and March 2, not exactly “today.” For another, of the six polls, only two are of likely voters, three are of registered voters (some 35-40 percent of whom will not vote) and one is of “adults” and thus totally useless to help forecast an election outcome today or any day (although such polls have a place).

    Applying a bit of political judgment Romney would very likely prevail today, although Virginia remains somewhat competitive.

    I think you will find that the many folks who construct similar maps do not rely simply on converting the latest polls into wins and losses.

  • Chase

    If gas goes to $6 a gallon, I’d bet on Romney.

  • nadine

    I’ll add one more point to John Burke’s: the 48%-44% totals you have vs. Romney leave 8% undecided. They will vote come November, and assuming no serious third party candidate, all things being equal, undecideds break against the incumbent.

  • Lorenzo from Oz

    Assuming war with Iran does not break out between now and November, it is going to be The Economy Rules. If economic recovery continues, the bread and peace model suggests Obama will be re-elected.

  • Charles R. Williams

    I can tell you that Kasich and Obama are equally unpopular in Ohio. The governor is unpopular because he is making the tough choices that are necessary to revive the state and, more importantly, because he lacks the political skill to remain popular as he does it. Obama and Kasich will not both be on the ballot in 2012.

  • Bart Hall (Kansas, USA)

    It would be a helpful tool to include the electoral votes in the rollover-popups.

  • Always On Watch

    Linked at the end of my post today.

  • Damir Marusic

    @nadine, if you look carefully at WRM’s post, we’ve gone and just distributed the undecideds 50-50.

    We’ve derived this from some quick and dirty back of the envelope guesses complex mathematical formulae using our state of the art computer facilities, including assigning undecideds to the two candidates (the poll numbers in the visual show the original RCP data without adjustments).

  • Damir Marusic

    @Bart Hall, we’re considering a separate page which will have a table of states below it, with electoral votes and percentages.

  • John Burke

    At the risk of becoming tiresome, the current state of the race simply does not support the conclusion that Obama would win “handily” if the election were held today. The way in which states’ electoral votes have been assigned in the map is misleading. Many other people have developed similar maps. The one to be found at the link has its limitations too but portrays the current situation more sensibly. It shows a much closer electoral college race, and paints a clear picture of where the battlegrounds are by painting the states as weak, strong or solid in their current partisan

  • Molon Labe

    Your analysis “[does not tally with my views].” It is so bad I am prepared to betb you that your Romney best case senario for the GOP is so bad that I’ll bet you 1,000 that your wrong. I much prefer people who aren’t so daft as to posit that Obama’s increase/decrease in popularity is uniform across the nation. Let’s face it, Obama isn’t losing popularity in California, nor is he gaining in Texas. The question is is he going to take VA, NC, PA, OH. Not likely. Butr will he lose Florida, Colorado, Nevada. Probably. In fact the question isn’t what states Obama gains, it what states will he lose. And the map shows Obama dropping like a rock. He is so bad that Satan no make that Nixon could beat him.

© The American Interest LLC 2005-2016 About Us Masthead Submissions Advertise Customer Service