Are Democrats Cannibalizing Election Day Votes?

One of the key distinguishing advantages to early voting is the ability of political parties to get their low-propensity voters to actually cast their ballot.  Every election, though, the question remains whether these turnouts in early voting are cannibalizing election day voters of each side.  The GOP ran a study on this concern and the results bode ill for what had been a key strength to Obama’s 2008 victory:

Democrats more than Republicans are getting their most loyal supporters to vote early, but with polls showing a close race among those who have voted so far, concerns are being raised about a GOP tsunami on the actual Election Day, next Tuesday.

According to a GOP analysis of early voting and absentee ballot requests provided to Secrets, the Democrats are turning out their most reliable, or so-called “high propensity voters” than Republicans, leaving fewer for Election Day. The GOP is pushing weaker supporters to vote early, expecting high enthusiasm to drive their regular supporters to the polls next week.

For example, in Ohio, the Democrats have turned out 43 percent of the most loyal supporters to vote, compared to just 27 percent of the GOP. In Iowa, the difference is 43 percent to 29 percent.

Even with the difference in turnout of loyal supporters, Gallup finds that among early voters, Romney is beating Obama 52 percent to 45 percent, though some state totals show an Obama advantage. Plus in states like North Carolina, Colorado and Iowa, the GOP has seen huge requests for last-minute absentee ballots this month.


  1. Brad
    Posted October 30, 2012 at 6:10 pm | Permalink | Reply

    You have to consider the source and you need more data. At this point, I’m not sure how much of this to believe.

    • Posted October 30, 2012 at 6:24 pm | Permalink | Reply

      I’d bet that the GOP has a lot of data on this. This is also not the first time I’ve heard this. Karl Rove and a few others have talked about it and a Democrat admitted to it, saying they were racking up the easy votes so they could get those who rarely vote to the polls on election day. This seems like a backwards strategy since it’d be easier to spread that out over a week or two as opposed to focusing on a massive election day GOTV effort.

      I tend to believe that they did it to hide the GOP advantage, knowing that the GOP turns out better than they do on election day. I believe McCain won the election day vote in 2008. That will increase this year and if Romney is leading the early voting by a decent margin then this could open way up next Tuesday,

    • William Jefferson Jr.
      Posted October 30, 2012 at 7:50 pm | Permalink | Reply

      “You have to consider the source.” The GOP is a suspect source?

      • wmart
        Posted October 30, 2012 at 8:40 pm | Permalink

        Yes…you never just take anything you hear from either political party at face value. I don’t just accept that Obama’s forays into PA and MI reflect no actual uncertainty because Messina and Axelrod told me to.

    • Brad
      Posted October 30, 2012 at 8:23 pm | Permalink | Reply

      Well, if Dem turnout is frontloaded it will start falling off in the next few days. So I guess we’ll see.

  2. Blackcloud
    Posted October 30, 2012 at 6:11 pm | Permalink | Reply

    We’ll find out November 6. Which, thank goodness, is only a week away. At least Disney and LucasFilm did us all a favor by annoucing their shocking news and distracting us from the Sandy and election talk for a while.

  3. Pete
    Posted October 30, 2012 at 6:13 pm | Permalink | Reply

    No one knows what to do when a strategy built entirely on the “bandwagon effect” collapses.

  4. No Tribe
    Posted October 30, 2012 at 6:19 pm | Permalink | Reply

    This you should enjoy Keith:

  5. No Tribe
    Posted October 30, 2012 at 6:25 pm | Permalink | Reply

    Things look terrible for Dems in NC, check out the charts:

  6. kenberthiaume
    Posted October 30, 2012 at 6:36 pm | Permalink | Reply

    new surveyusa ohio poll, o is up 3.

    weird though their party breakdowns are 38/32/26 adn they don’t add up to 100. adds up to 96.

    then if you do the math it doesn’t come out to o +3.

    They haev it R: 90/6, d 8/88 and ind: 48/37 comes out to about o + .6.

    If you use 41/32/27 it comes out to obama up 3.

    So once again using a pretty heavy D+9 poll. I wonder if they weight by party or just by who answers the phone.

    Also they say that 25% “have voted already”, but the site I’m looking at says as of 10/27 only 17% have voted already. They have another 15% or so who WILL vote and then Romney is ahead 50-41 amongst people voting on election day.

    So once again, kind of disappointing in Ohio. But it’s hard to know where they’re getting their party weights from.

    • Bryan
      Posted October 30, 2012 at 6:46 pm | Permalink | Reply

      Of the people they surveyed, 49% reported voting for Obama in 2008, and 40% for McCain. Obama won Ohio by only 4.6%. Looks like they oversampled Obama voters.

    • Eric
      Posted October 30, 2012 at 6:46 pm | Permalink | Reply

      Look at the sample in the poll. Almost all registered voters were determined to be likely voters. That means that it’s practically a poll of likely voters. Only 65-70% of registered voters actually vote. Survey USA has leaned about 4-5% to the left in every one of their polls for this reason. They are Registered Voter polls. RV polls are about 5% more favorable to Obama than LV polls.

    • JGS
      Posted October 30, 2012 at 6:47 pm | Permalink | Reply

      It has also been noted in SUSA’s prior OH polls that their relative weighting by geography leaves much to be desired. This poll has the following geographical weightings, would be curious as to how this compares to actual geographical allocations for 2004, 2008 and 2010:

      9% Toledo, 12% Dayton, 16% Cincinnati, 20% Columbus, 39% Cleveland, 4% SE Ohio.

      Their most recent OH poll was weighted:

      9% Toledo, 11% Dayton, 14% Cincinnati, 21% Columbus, 41% Cleveland, and 4% SE Ohio so pretty similar. My recollection is that there was some criticism, when that prior poll was released, about those weightings.

      • kenberthiaume
        Posted October 30, 2012 at 6:52 pm | Permalink

        national review said that cleveland area had decreased in population by 5% whereas the suburbs of cincy and dayton had grown in the past 4 years. CNN has cleveland plus “NE Ohio” at 37% from their exit poll in 2008.

    • TheTorch
      Posted October 30, 2012 at 6:55 pm | Permalink | Reply

      You really should just ignore these polls for Ohio from the left leaning pollsters, they are doing precisely what I stated in a posting a few days ago. This is really the last throw of the dice for Obama, they will do their best to prop him up, and the way to do that is show Ohio tight, or Obama with a lead. This allows the media to run with the narrative they want that Ohio is tight or Obama has a slight edge.
      Simple fact is if Obama has a lead of 3 in a D+9 poll he is losing Ohio, and probably by more than a few points.

  7. kenberthiaume
    Posted October 30, 2012 at 6:49 pm | Permalink | Reply

    Also I left out a key thing about the survey USA poll. They pose the question: “Who did you vote for in 2008?”. In that question Obama wins by 9, 49-40. Whereas he only one by 4.6 or so. So they’re kind of spotting him a few points, unless their contention is that many republicans from last time aren’t voting, or that many new democrats moved in from out of state or republicans moved out…just seems a little strange. How does it not seem strange to them? Last poll they had obama winning by 11 in 2008 with their people who had voted before.

    • Posted October 30, 2012 at 6:57 pm | Permalink | Reply

      “They pose the question: “Who did you vote for in 2008?”. In that question Obama wins by 9, 49-40. Whereas he only one by 4.6 or so.”
      That sounds to me like they talked to far more Democrats than Republicans (I’ll try to feign shock here).
      ~ Brittany

    • Eric
      Posted October 30, 2012 at 7:11 pm | Permalink | Reply

      It’s because they aren’t screening their voters. A lot of the people they are talking to didn’t vote in 2008 but would have voted for Obama if they had found time or whatever their excuse was. A month of early voting and there’s no time. They say they will vote this time, so Survey USA includes them in the sample. That’s why the sample is off.

      This isn’t new either. They do this in every election. It’s called a registered voter poll. Fact is that most non-voters say they are Democrats. Include them in the sample and Voila! you make the sample D+6 when Ohio is actually balanced politically. Most polls are doing this. Don’t let it stress you out. If they had Ohio in the bag like they claim, then why are they running ads in Michigan, Pennsylvania, and Minnesota? Why is Bill Clinton in Minnesota? Why is Biden in Pennsylvania? Why is the Romney campaign launching ads into Pennsylvania?

      • Blackcloud
        Posted October 30, 2012 at 7:33 pm | Permalink

        Polling 101: Respondents always indicate they voted for the winner in the previous election at rates higher than the winner’s vote share. Why? Simple: who wants to say they voted for a loser? Other reasons the “who did you vote for last time” number is usually off are that people don’t remember who they voted for or those who didn’t vote say they did. This is a perennial flaw in polling, but it’s not a structural one; it’s squarely a psychological one on the part of the respondents.

  8. Chris
    Posted October 30, 2012 at 7:37 pm | Permalink | Reply

    Maybe I’m missing something, but how does the GOP know which voters the Dems are turning out? Are they basing that on where the votes are coming from?

    • Ryan J.
      Posted October 30, 2012 at 8:39 pm | Permalink | Reply

      They are doing data mining. They will get copies of the voting rolls and look at where they live, how many times they’ve voted, if they are high propensity voters, things they may buy, and all sorts of other pieces of data. Then then they put it all together and have a pretty close idea how that person votes. Once that person votes that’s how they get their total figures of high and low propensity voters.

  9. Andrew
    Posted October 30, 2012 at 9:44 pm | Permalink | Reply

    What difference does it make either way? A vote is a vote.

  10. christopher
    Posted October 31, 2012 at 3:34 am | Permalink | Reply

    Just found this tidbit. I’ll leave the analysis to those more capable on this website. But I certainly like the margin and trend. I do know that these three papers are 2,3 and 4 in terms of size in the Florida marketplace…. 6 days and counting.

    A new poll shows Mitt Romney with a seven point lead over President Obama in Florida. The poll by the Miami Herald, El Nuevo Herald, and Tampa Bay Times shows Romney with 51% percent to Obamas 44% with a 3.5% margin of error.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: