Princeton Election Consortium

A first draft of electoral history. Since 2004

Early voting picks up steam

October 12th, 2014, 8:48pm by Sam Wang


Early voting has started, most notably in Iowa, Florida, and North Carolina. Here is a rundown by Michael McDonald, who drills into the subject in amazing detail. While we’re at it, here’s his early voting tabulation page. Bookmark it!

P.S. For general comments use the MSNBC thread. There’s a great conversation going on there.

Tags: 2014 Election

30 Comments so far ↓

  • Joe

    Sam, more questions and insight about CO SUSA poll. Seems they did two in CO one right after the other, the other being for the Denver Post. 2 point swing in D direction. The 2 points insignificant (seems within MOE), noise or perhaps the beginning of a swing back to Dems? I’d be suspect if they were at the same time, but they conducted 2nd one after the 4 point one.

  • J

    One of the things I am interested to know, and I hope it can be answered, is how much of an effect can ballot measures and/or the governor contests in the states with close Senate races have on those races? Could an overwhelmingly popular governor (or senator for that matter( potentially swing a close race in that state in a similar fashion to the Presidential coattail effect? Or, in the case of a high-profile ballot measure that drives out voters who would have otherwise stayed home, impact a race either way? Thanks.

  • Joe

    Sam, what are your thoughts on the SUSA poll out of CO today? It’s been argued by a few that the crosstabs are woefully wrong i.e. the hispanic vote totals and sampling size. would like to hear your take.

    • Sam Wang

      I think internals of a poll are good for examining a particular demographic’s views – they’re like exit polls that way. However, I do not think it is ever profitable to pick over internals as a means of questioning the overall result. In psychology, that is called motivated reasoning, i.e. you look harder when you don’t like the result.

      In short, the Survey USA poll was (a) unfavorable to Udall, and (b) in line with other recent results. This new poll confirms other recent data from Colorado. The next poll can always turn up a surprise, of course.

  • A New Jersey Farmer

    Be careful about the early voting numbers. I remember in 2012 that the early requests were overwhelmingly in the Republicans’ favor in Ohio and we know (happily) how that turned out. The news is good, but I’d like to see some polling that reflects Democrats coming out in larger numbers.

  • Aaron Booth

    So the current probability for D+I is 30%?

  • CRM

    I also see that in the brand new DMR poll there was a TWENTY point lead of 58-38 for Braley in early returns. Independents breaking again Ernst, too.

  • Dean

    I saw the early voting breakdown in a state that has suddenly become very competitive for the Democratic Senate candidate: Georgia. There is no partisan breakdown, but there is a noticeable decrease in white early voters, and there is an increase in new registration voters.

    There are some good early voting numbers for Democrats also in Iowa and North Carolina.

    GOTV efforts and early voting could mean the world to Democrats in the quest to keep the Senate, or to keep the Senate out of Republican hands via Democrat+Independent seats.

    In Illinois, my state, voting is promoted via early voting and via a ballot measure that will ensure voting rights. This is the opposite of what’s happening in some states, in which Republicans enacted bogus and illegal voter ID laws, whose sole purpose is to disenfranchise and reduce Democratic voters.

    The Democratic Illinois Senate candidate is trouncing his opponent in the polls, but GOTV may play a huge role in the gubernatorial race, in which Democrat Pat Quinn has union support and a large GOTV infrastructure. Challenger Bruce Rauner, who made $60 million last year after having stepped down from his company in 2012, has googobs of money to throw at a GOTV effort.

    This all is so fascinating, and due to the closeness of some key races, many of us are at the edge of our seats.

  • Canadian fan

    If Democrats hold North Carolina, Colorado, and Iowa, and Orman wins Kansas ( and decides to caucus with Democrats ), the Democrats will retain the majority even if they do not pick up any additional seats. So they can lose Alaska, Arkansas, and Louisiana, and still retain the majority. But I also think that all these races are much closer than people think. Georgia has great potential ( if all the votes are counted ). And South Dakota may be the new Kansas ( especially as Pressler might very well caucus with Democrats in the event that Weiland didn’t win ). Kentucky could be a surprise and unanticipated upset. In short, Democrats have a very plausible path to 50, and enough potential leg room to expand beyond that.

    • Joe

      Agreed. I think the beltway press knows that too, cause every time an article from a blog or new media seem to point out the real GOTV advantage Dems have been attempting to build money and body wise, Politico, The Hill, First Read and their ilk put out a “GOP faces easy path, or better odds, or Dems Panic” type story. GOTV is key, and some of the Iowa early numbers are starting to give me hope that IA stays blue.

    • Matthew

      You sure about that? Here is what Michael McDonald writing in the Huffington Post yesterday said about Iowa “As of Friday, Democrats lead Republicans by 22,352 total ballot requests 122,745 to 100,393, or 44 percent to 36 percent. At the same number of days from the election in 2010, Democrats led by 27,759 requests or 49 percent to 35 percent. A 14 point margin in 2010 shrank to 8 points.” So in terms of percentages, Democrats are down five and Republicans are up 1. And in straight ballot terms the Democrats margin is lower by 5,407 ballots, as of last Friday.

    • Davey

      Matthew – that’s some “fuzzy math” that I’m not sure indicates much either way. We don’t know whether the growth in early voting increases turnout…are these new early voters people who are new to midterms, or they just don’t want to go the polls like they have in the past. Regardless, the “no party” share exploded relative to the others, so it’s impossible to discern much until we have election result data. It will be a fun year to compare 2010 to 2014 once the fights are all settled and people aren’t so dang hyper.

  • RAJ

    The Real politics poll aggregator has the Democrats with 46 safeseats and the GOP with 45 safe states. There are 9 tossup states, 6 Democratic seats and 3 GOP seats (not counting South Dakota now a tossup state). To get to 51 seats the GOP has to win all Democratic held seats and lose none of the GOP held seats. If the GOP loses one of the nine they can’t get to 51. The GOP is drawing to an inside straight flush whose odds are 0.08.

    http://www.realclearpolitics.com/epolls/2014/senate/2014_elections_senate_map.html

    • JayBoy2k

      Raj,
      While your overall point has some merit, your calculation is flawed because 2 Senate seats (Montana & West Virginia) are currently held by Ds and are as sure to switch as states like Michigan and Minnesota are sure to stay Ds.
      So, Republicans need only 7 of the 9 and currently lead in 7 of 9 polls.
      That is why many call this a coin flip rather than a long shot either way.

  • axt113

    I think there is an issue with your code, the probability line at the top of the page for the Dem+Ind on election day says 0.4

  • Canadian fan

    A wonderful article on the breakdown of the early polls. I agree with atothec. It’s an absolute disgrace that 50,000 votes might be uncounted. As October 6 was the final date for registration in Georgia, this election will likely hinge on whether those votes are included or not. This race could very well be the Florida 2000 election of the cycle – headed for litigation – big time. Even with Florida 2000 in mind, however, it’s hard to imagine 50,000 votes being brazenly uncounted. In Florida there was of course the famous butterfly ballot where people intending to vote for Al Gore ending up voting for Pat Buchannan. That was thousands of votes right there, and not only would have changed Florida’s results, but subsequently history.

    However, I still see the Democrats path to victory as being North Carolina, Colorado, and Iowa, under the assumption that Orman carries Kansas and decides to caucus with Democrats. In that scenario, Democrats could lose Alaska, Arkansas, and Louisiana, not pick up and any seats, and still retain the majority.

    Having said that, I think all of these races are within Democrats’ grasp, including Georgia – if the votes are counted. And South Dakota looks very possible, even if Pressler wins, as it is quite possible he could caucus with Democrats.

    But with each election cycle there is generally something that nobody anticipates. I think Kentucky will be that race.

  • Olav Grinde

    So here is a simple question: Are there any large-sample, state-level polls specifically asking early voters what their choices are?

    • Sam Wang

      See the Des Moines Register poll. Braley by 15-20 points IIRC. Also see McDonald’s analysis.

    • Olav Grinde

      Thank you, Dr Wang.

      In Prof. Michael McDonald’s overview, I find it striking 25 % of early Georgia voters did not cast votes in the 2010 election. For North Carolina the figure is 31 %.

      I realize that the numbers are still modest, but those two figures give me hope that we’ll be in for a positive surprise.

  • JayBoy2k

    There is a very good article from the Hill today. Quotes Sam a couple of times and tells us that some believe that Nate has been taken over by the Koch brothers!! and then this insight..

    “The Democratic Congressional Campaign Committee sent out a fundraising email early Friday afternoon with the subject-line, “New poll proven WRONG.” The poll in question, from Rasmussen Reports, showed Democrats up two points in the race for Congress.”
    http://thehill.com/homenews/campaign-polls/220435-dems-dont-trust-the-polls

    • Sam Wang

      Proven “WRONG”? Before the election? That is quite a trick. Also…I am not following why the DCCC is against a poll showing D+2%, which is favorable to them, not far from the current median.

    • Katie

      I’ve ended up on the DCCC mailing list a few times. Most of it got sent to my spam folder anyways. Basically, they try to make every minor thing into a chance to scream that the election is doomed and unless they get $50 from you the Republicans will take over everything forever. Most subject lines involve things along the line of “emergency” or “last chance to save us”.

      They could take the reasonable approach of letting people know that just because one poll shows they’re ahead doesn’t mean that the election is called, and that they still need donations to build momentum and a safe margin, but that’s not alarmist enough for whoever writes their emails.

  • atothec

    Ugh this is the new meme this cycle. Never mind that in the battleground states the high level of engagement is obvious by the close races in what should be republican cakewalk states.

    It’s also a slap in the face to the 100,000 new voters in Georgia, half of whose votes are being denied by the SoS. So it’s bizarre to see this making the rounds with all that going, including the Iowa early returns which show a high turnout, again especially due to new voters.

    The ground game is real and it can win the Senate. The issue is not whether Dems will turn out (in the battleground states) it’s will their votes be counted.

    • Olav Grinde

      It’s also a slap in the face to the 100,000 new voters in Georgia, half of whose votes are being denied by the SoS.

      Democrats should loudly call this by its proper name: Voter fraud!

      This, like so many other Republican measures, aim to deny the right to demographic groups that tend to vote Democratic.

      It is the only significant voter fraud America is seeing.

    • Joe

      Actually, it’s 210,000 that have been processed, with another 40,000 that have “vanished”. Those are the ones that have been complained about. Besides the point though, I live here in GA and I can tell you we are engaged and ready to go! GOTV!

  • A New Jersey Farmer

    Interesting article and poll that says that the public wants a Democratic-led Congress, but that Democrats are less interested than Republicans in actually voting. We’ve discussed the ground game. It is time to get involved.

    http://blogs.wsj.com/washwire/2014/10/12/poll-democratic-supporters-less-interested-in-midterm/?mod=WSJ_hpp_MIDDLENexttoWhatsNewsForth