MarkGrebner.Org

Straws in the wind – the early voting in Michigan

—

by

Violet
in
by: Grebner

Thu Jan 03, 2008 at 23:58:38 PM EST

I have just looked at a file of absentee ballots requested and returned in Michigan for the January 15 election.  Although absentee voters aren’t completely representative, I work with the data I have, rather than the data I’d like to have.  (Apologies to former Secretary Rumsfeld.)

 UPDATE – 1-15-08 – Been looking at actual counts of ballots returned by municipality, and I think I’ve been guessing too high a turnout.  The Democratic percentage remains stubbornly high, but I’m beginning to think the Republican turnout won’t be as high as 900,000.

 It’s foolish to update my guess, because I always discover after the election that my best guess was made 4 weeks out, and my last-minute adjustments are always in the wrong direction.  But I can’t help myself.  

So, my final guess:  750,000 Republicans, 600,000 Dems.  Wrong, wrong, wrong.

Not all local Clerks use the State’s system for handling absentee ballots, so my data only reflects about 75% of the population.   As of today, they’ve issued 260,000 ballots, and received about 120,000 returns.  Guessing how many will be requested and returned in the remaining twelve days, and assuming 20% of the ballots are in areas not covered by my file, I estimate there will be about 350,000 absentee ballots cast.  If absentees make up 25% of the total, we would expect 1.4 million total voters, slightly below my previous guesses.

 More interestingly, the partisan distribution is much more Democratic than I would have guessed.  On my firm’s master file, we have coded the average registered voter in Michigan about 52% Democratic, while the people requesting ballots are about 51% Democratic – virtually the same.

 I think some of those Democrats are casting ballots in the Republican primary, but not very many, at least among absentee voters.  On one hand, the lack of competition and the effects of the DNC sanctions must be discouraging Democratic turnout, but that reduction must be offset by the greater enthusiasm and more interesting race on the Democratic side.

Counting the Democratic crossover, I guess about 55 or 57% of the turnout will be in the Republican primary. 

 I will post updates to reflect additional information as I discover it.

Comments

19 responses to “Straws in the wind – the early voting in Michigan”

  1. Violet Avatar
    Violet

    Michigan Democratic Turnout
    So, 1.4 million total voters, 44% are Democrats, equals 616,000 Democratic voters. It’s a wise move on greee’s part that he didn’t take the bet and challenge my prediction two months ago that over 500,000 people would vote in Michigan’s Democratic Primary. If 1.4 million people do vote in the Presidential primary, that would outpace both 1992 and 2000. We haven’t had a Presidential primary that well attended in over 30 years. Apparently, voters aren’t feeling as disenfranchised as certain bloggers.
    by: northernlib @ Thu Jan 03, 2008 at 23:44:34 PM CST

    Reply
    1. Violet Avatar
      Violet

      Confusing terms.
      “616,000 Democratic voters”, meaning Democrats who vote in their own (meaningless) primary. Plus roughly 100,000 more who vote in the Republican primary. That is, when/if the exit pollers ask people leaving the polls which party they identify with, they may get more D than R answers.
      Of course, this is all based on absentee voting, and possibly the pattern will be somewhat different among walk-ins. Maybe there will be a higher crossover rate, because (unlike some of the absentees) by election day all the voters will know the Democratic ballot is missing some of the major choices.

      by: Grebner @ Fri Jan 04, 2008 at 03:03:51 AM CST

      Reply
      1. Violet Avatar
        Violet

        If by this:
        …by election day all the voters will know the Democratic ballot is missing some of the major choices.
        you mean, after they get in the booth and SEE that their guy is not on the ballot – but, that is evidently not what you mean, since you are predicting that their knowledge will change which ballot they select.

        On the contrary, I predict that a significant number of Democratic voters will get to the booth blissfully unaware that their guy (Obama, in most cases) is not there. I base this on conversations I am having right now, 11 days out, with students and others who have no clue and fully intend to go to the polls and vote for Obama. When I tell them otherwise, they look at me dubiously, too polite to disagree with their professor, but I can tell they think I MUST be mistaken.

        What is going to change in the next ten days?

        And – if I’m right – how to quantify the disillusionment, discouragement, disgust, and demoralization?

        by: memiller @ Fri Jan 04, 2008 at 09:46:46 AM CST

        Reply
        1. Violet Avatar
          Violet

          Remind them
          You can remind them that their candidate felt the votes of their peers in Iowa and New Hampshire were more important than the votes in Michigan. They always have the option of voting for Clinton, Kucinich, Uncommitted or causing havock on the dark side.
          by: northernlib @ Fri Jan 04, 2008 at 09:59:42 AM CST

          Reply
          1. Violet Avatar
            Violet

            So that truce
            lasted all of half a day…
            So obvious…but at least your open and honest about it now.

            by: Nazgul35 @ Fri Jan 04, 2008 at 10:19:50 AM CST
            [ Parent ]
            You just don’t get it. (4.00 / 1)
            I am reporting facts on the ground – and your reply is that I could take the opposite side of the argument (which we have hashed out ad nauseum here) than what I actually hold.
            It’s not about what I think or what you think. It’s not about “reminding” these voters about anything, since (unlike us) they are totally unaware of the backstory of why the candidates withdrew in the first place. If you read my comment, they are totally unaware that the candidates HAVE withdrawn.

            Arguing that a group of voters “should” think differently, or “should” be better informed, is meaningless. Facts are facts. If you have any evidence that significant numbers of Michigan voters are NOT unaware of the ballot situation, report that.

            by: memiller @ Fri Jan 04, 2008 at 10:43:06 AM CST

          2. Violet Avatar
            Violet

            Totally Unaware
            Well, they apparently aren’t totally unaware because you have seen fit to educate them about it. It’s not like there has been a lack of media attention to this issue in Michigan. Quite honestly, if someone is totally unaware of the primary situation, how aware are they about any of the other issues facing our state and nation? Do we really want our elections to be just about voting for a certain candidate because it’s the trendy thing to do? Quite frankly, along with the Democratic and Republican ballots, there ought to be a third option labled “Ignorant” for the whiners. It can list Brittany, Paris and Zac Efron.
            by: northernlib @ Fri Jan 04, 2008 at 11:12:19 AM CST

  2. Violet Avatar
    Violet

    That’s an awful big guess
    As of today, they’ve issued 260,000 ballots, and received about 120,000 returns. Guessing how many will be requested and returned in the remaining twelve days, and assuming 20% of the ballots are in areas not covered by my file, I estimate there will be about 350,000 absentee ballots cast.
    Could you be explicit about how you come up with a number of cast ballots that is 90,000 higher than the number actually requested with 12 days left?

    Thanks…

    PS: Still waiting for you to post online the raw data you offered to supply to test the validity of your results….

    by: Nazgul35 @ Fri Jan 04, 2008 at 10:26:17 AM CST

    Reply
    1. Violet Avatar
      Violet

      Grebner
      Grebner doesn’t need to explain his methods. Grebner just knows.
      by: northernlib @ Fri Jan 04, 2008 at 11:28:32 AM CST

      Reply
      1. Violet Avatar
        Violet

        Actually he does
        and if you knew anything about statistics, you would be somewhat alarmed at some of his posting on this site re: internal validity, ecological fallacies, etc.
        by: Nazgul35 @ Fri Jan 04, 2008 at 12:48:17 PM CST

        Reply
        1. Violet Avatar
          Violet

          Estimation.
          First, I expect almost 90% of the ballots requested to be returned – that’s typical.
          Second, because the process was so delayed by – ahem, somebody’s lawsuit – the process of mailing applications and ballots was delayed generally, so I think there are still large numbers going out. It’s hard to extrapolate with any assurance, because recent numbers have been affected by the holidays and the dates various courthouses were open to receive mail. But on Wednesday, the collective Clerks of Michigan mailed out 18,000 ballots, which must have been assembled by crews working over the holiday weekend. On Thursday, they had mailed almost 1000 ballots by the time my copy of the file was made (about noon). And they had already logged 2000 more as going out today (Friday), presumably meaning they had received the requests and were assembling the outgoing packages but expected to miss Thursday’s mail. Like I said, it’s hard to draw a line through those points, but it seems as if they’re still shoveling ballots out about as fast as they can, and I would think some voters have put off dealing with the requests until after the holidays.

          Finally, as I replied to your earlier posts, I have no idea what you want. Your only responses have been equally cryptic.

          by: Grebner @ Fri Jan 04, 2008 at 12:19:57 PM CST

          Reply
          1. Violet Avatar
            Violet

            Dude,
            come on. I was very clear and will quote my comment for the third time here, because it seems like you aren’t clicking through the links…
            YOU

            I’d be happy to supply the data
            If you send me your email (I’m “Mark dot Grebner at Gmail dot com”) I’ll mail you a cd with the data for everybody in your precinct. This offer applies to everybody – by the way – who’d like to do such a comparison.
            I’ve spent 35 years – yes, since the days of punched cards – compiling voter lists. I know something about the trade.

            ME

            You don’t need to go through all this.
            A very simple solution is for you to select a precinct (your choice) that you don’t have to identify and post it online along with an outline of your model and the tests you conducted in determining the specification of the model.

            You can load these up as files that are down loadable…your choice (excel, sass, stata, spss, minitab, tab or comma delimited data, etc).

            With that it should be easy to verify your claims about the accuracy of your predictions…

            YOU

            Methodological/theological disputes
            I don’t believe a model’s validity can be checked or judged by examining it for internal consistency. The only valid test is whether it predicts new data better than competing models. (Or data that is at least “new” to the models being examined.) As it happens, my current model of partisanship has about fifty terms, many of which involve lookups into related tables, and some of which are multiplicative or compound. In other words, to understand my model, first you’d have know FoxPro, and then you’d need to set up trace/debug to see how the program executes. At some point (once I have caught up on sleep) I plan to post the general scheme to my Technical Politics diary. It’ll probably take 5000 words to cover the basics.

            ME

            All it requires
            is for you to provide the data, as you suggested before in a previous post. You can provide a pre-test data set with a post test result test for the precinct. What won’t be able to be assessed is whether the alternative I suggested would be any better because we don’t have the data. It won’t be “difficult” to anyone to assess your methods who has more than a passing understanding of stats and can use sophisticated testing techniques.

            ME…still waiting for the data.

            by: Nazgul35 @ Fri Jan 04, 2008 at 12:58:32 PM CST

  3. Violet Avatar
    Violet

    You have perfectly reprised our previous exchange.
    And I still have no clue what you’re talking about.
    by: Grebner @ Fri Jan 04, 2008 at 15:04:42 PM CST

    Reply
    1. Violet Avatar
      Violet

      Good God man
      Take some data from your vaunted files that you say are sooo damn good…
      Put them in a downloadable format with a description of your methods used…

      So that the validity of your data and methods can be assessed.

      If you don’t understand the statistic or internet jargon, get someone to translate it for you…

      Or are you just stalling?

      Geez.

      by: Nazgul35 @ Sat Jan 05, 2008 at 15:15:59 PM CST

      Reply
      1. Violet Avatar
        Violet

        Did I see, at some point, that he provided you with an e-mail addy?
        Yes, yes, he did. He also asked for your e-mail address.
        I’m not saying you should stop demanding information from him. What I am suggesting is that when he provides an avenue for you to inquire directly, that it would be — to me — preferable that you do that rather than hector him in a way that prompts people not involved in the conversation to e-mail me and ask me to make it stop.

        Among the Trees

        by: Eric B. @ Sun Jan 06, 2008 at 23:15:03 PM CST

        Reply
        1. Violet Avatar
          Violet

          Wrong.
          One thing you need to understand about science is the need for replication of results. If it was just me, it would quickly turn into he said he said…
          The method I put forward would allow for more than just myself to analyze his “data” methods…which is much fairer.

          It would be easy for me to create an email address and get this data, but I am for full disclosure and transparency with regards someone who influences the Michigan Party so heavily.

          It should not be just me doing this.

          He can do this in a manner that protects his proprietary status re: data, but that allows for the analysis of said data using standard scientific inquiry.

          As for the people asking you to make it stop, what is their interest in protecting Mr. Grebner from an open and honest analysis of his claims re: his data?

          Mine is scientific…what’s theirs?

          by: Nazgul35 @ Mon Jan 07, 2008 at 11:46:29 AM CST

          Reply
          1. Violet Avatar
            Violet

            I am going to repeat myself…
            Did I see, at some point, that he provided you with an e-mail addy?
            Yes, yes, he did. He also asked for your e-mail address.
            I’m not saying you should stop demanding information from him. What I am suggesting is that when he provides an avenue for you to inquire directly, that it would be — to me — preferable that you do that rather than hector him in a way that prompts people not involved in the conversation to e-mail me and ask me to make it stop.

            Do you know what happens every time I get a complaint about content? The first thing is that it usually requires that I spend about half an hour trying to figure out what the hell the person is talking about. The second is that it probably takes another half an hour to sort out the problem and figure out where everyone is coming from. The third is that it takes another half an hour to arrive at some kind of conclusion and word a post, comment, or response. That means, it usually takes about an hour and a half of my time, time I spend doing nothing else but dealing with a complaint and that is eaten out of my personal life or that which I could have used to write something (this, for instance, comes rather than writing a blog post I’ve been researching for three days).
            In this case, you could satisfy everyone — yourself and Mark Grebner included — by simply using the e-mail address he provided to contact him directly. Then, you could write a diary about it, and it would be front paged.

            Among the Trees

            by: Eric B. @ Mon Jan 07, 2008 at 12:46:49 PM CST

  4. Violet Avatar
    Violet

    Analyzing Complaints
    I’d be happy to help you with this in the future Eric. As you know, I have a tendency to be impartial in these types of matters. I figure one of these days, the gods will see fit to give me the magic ability to promote diaries and delete posts.
    by: northernlib @ Mon Jan 07, 2008 at 13:26:05 PM CST

    sorry
    about the people who apparently aren’t willing to come on the blog and post their concerns, and you having to deal with it…
    But I am going to continue to Ask Mr. Grebner to post data online in a manner that will allow for anyone who has the ability to test his data for accuracy and validity.

    Given his comments on this site, I have grave concerns about what he is telling people his data does and what it actually does.

    Why are Mr. Grebner and his fans afraid of simply exposing his data and methods to the light of day?

    It couldn’t be because he has a financial interest that his work appear to be scientific.

    The truth is, from what he has posted on this site, his scientific rigor appears to be shoddy at best. He posts diaries on this website that are taken as if they are scientific.

    That is my objection. If he wants to portray himself as a pundit or political operative, I have no qualms with that.

    I am even enough of a scientist to allow that Mr. Grebner’s methods and data are first rate…if that is what I find.

    I have no agenda other than I want Mr. Grebner (and any other individual who purports to conduct scientific inquiry) to hold to the very simple standard that we have in science…and clearly Mr. Grebner has stated that he doesn’t have to. Everyone should be aware that many things Mr. Grebner engages in violate the ethical standards of science that academics follow.

    You publish your data so that the results can be tested by a large number of individuals, who have the skills to test for a variety of factors, and assess the claims you are making.

    It’s simple. If Mr. Grebner wants to forgo such scrutiny, than stop engaging in psedo-science.

    However, if he is interested in scientific inquiry, he should have no qualms about publishing his methods and data to scrutiny from the large scientific community…something we are required to do in academia to keep our jobs.

    by: Nazgul35 @ Mon Jan 07, 2008 at 14:46:07 PM CST

    Reply
  5. Violet Avatar
    Violet

    Michigan Democrats
    Due to violating party rules, the Democratic National Committee has stripped Michigan of all its delegates to the Democratic National Convention. You can vote in the Democrat’s primary, but no Michigan delegates will be sent to the DNC to vote for you.
    As if that isn’t bad enough, most candidates running for the Democrat’s nomination didn’t even register to be on the ballot or have withdrawn their names. Here is a list of people who will be on the ballot: Hillary Clinton, Mike Gravel, and Dennis Kucinich. So if you wanted to vote for Barack Obama, John Edwards, or any other candidate you’re SOL because they won’t even be on the ballot.

    Instead of casting a meaningless vote on the Democrat’s ticket, you can vote in the Republican primary, helping bring a “win-win” ballot for the 2008 Presidential election by voting for Ron Paul. The war in Iraq has consistently rated the highest issue important to voters. Another issue not polled which I’m sure matters to voters is the restoration of our civil liberties.

    Ron Paul is the only Republican candidate who is against the Iraq war and for the restoration of our civil liberties. All of the Democrats oppose the war in Iraq, so by voting for Ron Paul in the Republican primary you can help create a “win-win” 2008 Presidential ballot. It will be an anti-war, pro civil liberties Democrat versus an anti-war, pro civil liberties Republican. No matter who wins the election, the war will end and our civil liberties will be restored.

    Visit http://www.RonPaul2008.com/Issues

    by: j6p @ Wed Jan 09, 2008 at 21:48:47 PM CST

    Reply
    1. Violet Avatar
      Violet

      UPDATE: I’m reducing my turnout guesses
      The Democratic percentage of the absentee vote continues to surprise me, but the total number of votes being cast by mail is weaker than I expected. See the updated diary text.
      by: Grebner @ Tue Jan 15, 2008 at 01:46:59 AM CST

      Reply

Leave a Reply

Your email address will not be published. Required fields are marked *