New CNN & Zogby polls contradict DMR: What's going on?

New polls today by CNN and Zogby both put Clinton in first place.  What are the numbers and how can you explain the difference between these and the DMR poll?  More below the fold

Here's the CNN results:
Clinton 33
Obama 31
Edwards 22

And Zogby
Clinton 30
Obama 26
Edwards 25

How can the discrepancies between, on the one hand, these two polls and a half dozen other recent polls, all of which show Clinton or Edwards on top, and the DMR poll, which shows a big lead for Obama?

Well, I think they are explained by what can only be characterized as the bizarre findings of the DMR poll.  Here was the turnout in 2004 Democratic Iowa caucus according to exit polls:

Iowa Democratic Caucus 2004
Total Turnout: 124,000
Democrats: 99,200 (80%)
Independents: 23,560 (19%)
Republicans: 1240 (1%)

Well, according to the DMR poll, the percentage of Independents and Republicans caucusing in the Democratic caucus this year will leap to 40% and 5%.  Projecting a turnout of 200,000 people, which would be a huge leap, this means numbers such as the following:

Iowa Democratic Caucus 2008
Projected Total Turnout: 200,0000
Democrats: 110,000 (55%) [number up 11% from 2004]
Independents: 80,000 (40%) [number up 340% from 2004]
Republicans: 10,000 (5%) [number up 806% from 2004]

Look at the increased numbers from 2004.  They are, to put it mildly, very unusual.

The DMR's projection is not impossible, but it seems to me very unlikely.

•  Bottom line?

  Iowa and NH need to go.  The last thing we need is Republicans and conservative Independents picking the candidate.  There is a diary just below this one examing the same thought.  Happy New Year.

  by dkmich on Tue Jan 01, 2008 at 05:11:45 AM PDT

  •  I read an analysis somewhere that

    this early states are so torn between three good candidates that they can't make up their minds.  I totally agree with you.  This system is completely for the birds.

    by DCDemocrat on Tue Jan 01, 2008 at 05:16:29 AM PDT

    •  Why Not

      Larger, more diverse early primary states that look like America?

      by bink on Tue Jan 01, 2008 at 05:42:44 AM PDT

      •  great signature, bink.

        by DCDemocrat on Tue Jan 01, 2008 at 06:07:39 AM PDT

      •  I love the micro-politics...

        ...that you get with the Iowa/NH start.  This forces the candidates to focus on feet on the ground organizations rather than media events and TeeVee.  

        Imagine if Florida and Pennsylvania were the first two primarys?   Hillary, Obama and Guiliani would have carpet bombed the major cities with their money while Edwards, Huckster and the rest would be dead in the water speaking to dozens at church socials.

        Winning Iowa or NH rarely dictates the nominees...it sets the stage and winnows out the weak.  Remember Pat Robertson or Pat Buchanon?

        BTW, great analysis of the polls and total proof WHY polls are full of sh!t.

        by neoeconomist on Tue Jan 01, 2008 at 06:28:28 AM PDT

  •  Sure...

    Let's allow the old-gaurd interests to dominate the party from here on out.

    by sunbode2000 on Tue Jan 01, 2008 at 05:16:37 AM PDT

  •  Who's Got Momentum?

    From the DMR's David Yepsen discussing his paper's poll:

    "the nightly results show Obama's support flat over the last two nights, Clinton's declining each night, and Edwards' support increasing each night."

    by flubber on Tue Jan 01, 2008 at 05:17:28 AM PDT

    •  Yet in Zogby

      which has a smaller standard error, Hillary remains stable and Edward declines.  Take your polls with a grain of salt.

      by DCDemocrat on Tue Jan 01, 2008 at 05:29:04 AM PDT

      •  Allergies

        Yet in Zogby which has a smaller standard error, Hillary remains stable and Edward declines.

        Huh.

        During the course of the three day Zogby tracker, Edwards is up a point, while both Obama and Clinton are down a point.

        But we all know you're pretty allergic to the truth, "DCDemocrat".  David Sirota has a wonderful diary about how we're coming to Washington to clear out the folks like you.

        by flubber on Tue Jan 01, 2008 at 05:35:07 AM PDT

        •  flubber

          today's Zogby is out, and Edwards is down one point while Hillary and Obama are static.  It's ridiculous to make anything out of a point, though.  Just saying.

          By the way, the last time you started this name calling thing, as I recall, you had a comment troll rated.  Let's stay away from name calling, flubber.

          by DCDemocrat on Tue Jan 01, 2008 at 05:40:09 AM PDT

        •  Going up or down a point in a poll...

          ...is too insignificant to warrant a mention.  

          Maybe if Edwards supporters would be less insulting to there brother Dems they wouldn't have to cling to the occasional one-point bounce polls.

          But we all know you're pretty allergic to the truth, "DCDemocrat".  David Sirota has a wonderful diary about how we're coming to Washington to clear out the folks like you.

          That's what I'm talkin about.

          by neoeconomist on Tue Jan 01, 2008 at 06:33:40 AM PDT

      •  Statistics ...

        Aren't even the issue. This is really a debate over who will show up, and that is probably the toughest problem in polling. Basically there are two ways to estimate likely caucusgoers, and both are fundamentally flawed.

        The subjective method is to ask people, the flaw being that people are liars, especially when asked about something they know they ought to do but probably won't. The objective method is to model your demographics on what happened last time, the flaw being that other people might show up this.

        What the hell is with Iowa? I thought it was full of nice people. Now we find out they are either liars, or the sort of people who crash someone else's party.

        by al Fubar on Tue Jan 01, 2008 at 07:45:36 AM PDT

  •  ludicrous

    We have ethanol proposed because of the Iowa primaries.  Buying  votes while screwing the country--so Republican.

    by melvynny on Tue Jan 01, 2008 at 05:25:01 AM PDT

    •  And ethanol is just a waste of corn.

      It is just another pay off to ag.   Band aids - don't you ever get sick of band aids?

      by dkmich on Tue Jan 01, 2008 at 06:50:40 AM PDT

      •  ah yes, that much less corn making diabetics

        If only the price of corn gets high enough to make hfcs too expensive to put in everything.

        by mollyd on Tue Jan 01, 2008 at 08:01:48 AM PDT

        •  There's a silver lining.

          And I'm serious.

          by Magenta on Tue Jan 01, 2008 at 08:32:13 AM PDT

        •  Poor people don't eat protein,

          it cost too much.  They eat starch and that would be corn.  The world doesn't begin and end in the USA.  

          Ethanol from hemp might be a better idea, but it  is still inadequate to the goal.  

          by dkmich on Tue Jan 01, 2008 at 08:43:14 AM PDT

•  DR poll has a better hold on their state

  I think they are seeing something that ONLY people on the ground have been seeing.
  But hey you never know. Read this article, it is a GREAT analysis of Iowa, stunning political reporting at its best.

  http://www.huffingtonpost.com/...

  by gaiilonfong on Tue Jan 01, 2008 at 05:13:53 AM PDT

  •  Not necessarily

    they predicted a significant win for Kerry both in Iowa (outside the margin of error) and nationally on the eve of the 2004 election.  And unfortunately we know how that turned out.

    The caucuses this year are very different to anything in the past on both sides of the fence, and noone is going to win by 7%.  Much more likely by 1% if even that.

    by Norwegian Chef on Tue Jan 01, 2008 at 05:18:28 AM PDT

    •  Polling caucuses are different than polling

      regular elections.  They are the most reliable to do the former.

      by Prince Georges for Obama on Tue Jan 01, 2008 at 05:20:05 AM PDT

    •  They still pick the winner of the Iowa Caucuses

      margin was bigger than they predicted, as for the general, I NEVER said they were good at that.
      Take a gander at this GREAT political reporting from the ground!

      http://www.huffingtonpost.com/...

      by gaiilonfong on Tue Jan 01, 2008 at 05:20:34 AM PDT

•  you hit the nail on the head

  it's all about the turnout models.  DMR is basing it's numbers on a very different turnout model than the historical one.  They could be right, but I wouldn't take it to the bank.

  As a Hillary supporter who thinks the best thing for her would be to win and the second best thing for her would be for Obama not to win in any case, I am not too worried.  Independents and Republicans won't pick the Democratic nominee, just as Independents didn't pick the Republican nominee in 2000 when McCain based his strategy on them.

  by markusd on Tue Jan 01, 2008 at 05:14:00 AM PDT

  •  Well,

    I am on a ground here and I have WITNESSED and HEARD Republicans and Indies who said they are switching their party on caucus night to stand for Obama.  And guess what, these are older folks who are FED UP and will never vote for Hillary.

    You guys don't get it.  To win in this country you must be appealing across the board.  Your message must be appealing across the board.  That is the main problem with the perception of Hillary Clinton.

    by icebergslim on Tue Jan 01, 2008 at 06:25:53 AM PDT

    •  me too.

      spent last night with two registered independent male neighbors here on my block in des moines who are coming out for obama...got an e-mail yesterday from a woman i'd been lobbying to caucus for obama. and she said that she'd settled on him, but most importantly, so had her husband--a registered republican...and several of her republican relatives in pella.

      i think there are a fair # of households like that here--wife is registered dem and husband is registered rep. so if you've got your wife to go with you, it doesn't make it as scary to come out and switch to dem for the night to participate in our process.

      again, this is all anecdotal. but i do believe that here in iowa--based on conversations with friends, family and neighbors--that there is a real interest in obama's candidacy on the part of independents and republicans...not saying they will definitely show up to caucus for him. but they are intrigued.

      by sunhaws on Tue Jan 01, 2008 at 06:45:25 AM PDT

    •  My point....

      Republicans picking the Dem candidate AND supporting Obama.  If you think that's good, you should join the Lieberman Party.

      by dkmich on Tue Jan 01, 2008 at 06:52:00 AM PDT

      •  Except Obama

        has one of the most progressive voting records in the Senate.  Clinton isn't even close.  Obama's appeal has as much to do with tone and rhetoric as policy.  If you are worried about someone siding with the Republicans too much, just ask yourself who did that more - Clinton or Obama? What were the results?

        by clambake on Tue Jan 01, 2008 at 07:03:47 AM PDT

        •  tone and rhetoric

          I don't like his tone or his rhetoric.  He and Hillary are "insiders".  I don't want an insider.

          by dkmich on Tue Jan 01, 2008 at 07:28:24 AM PDT

        •  clinton isn't even close?

          clinton and obama's voting records are extremely similar from any vaguely objective point of view.

          carry on.

          by nbutter on Tue Jan 01, 2008 at 03:12:52 PM PDT

  •  Sounds like you are hoping and rationalizing

    Clinton and Edwards are losing some support at the wrong time; Obama is gaining at the right time.  I don't think DM would make a statistical inference without a very strong degree of certainty.  Keep in mind that the weather is predicted to be clear and mild.  It will be sunny, no wind, with temps between 20 and 30.  

    The bottom line:  it is close, but Obama is in the driver's seat.  And America will be better for it.

    by daninoah on Tue Jan 01, 2008 at 06:36:47 AM PDT

•  Actually

  the CNN poll shows Hillary and Barack in a statistical tie (with both on an uptrend and Edwards on a downtrend) and Zogby shows a slight overnight trend in Barack's favor (and not in Clinton or Edwards favor). The polls are all over the place, but DMR is still the gold standard of caucus polling.

  by Prince Georges for Obama on Tue Jan 01, 2008 at 05:14:22 AM PDT

•  Easy

  I think that the figures from 2004 could easily be eclipsed by a factor of 2 or 3.

  Yes, I really believe that there could indeed be doubling of caucus-goers this year and that new-comers will represent a MAJORITY of caucus-goers.

  This is the most exciting race (possibly) ever. For every person that attended in 2004 there just has to be 2 or 3  new people that want to take part this time around.

  by sunbode2000 on Tue Jan 01, 2008 at 05:14:32 AM PDT

  •  3X is a stretch

    But approaching 2 X 2004 would not shock me - the highest prognostication I have seen is 200K, and that was from an enthusiastic blogger.  The highest MSM guess I have seen is around 170K.  I am not on the ground (though I do have some Iowa ties), so my assessment of the level of excitement comes from within the Obama campaign, but it seems that it is orders of magnitude greater than 2004.

    by clambake on Tue Jan 01, 2008 at 07:12:42 AM PDT

    •  OK

      just looked it up - the 200K I saw was actually from the DMR poll.

      by clambake on Tue Jan 01, 2008 at 07:17:40 AM PDT

•  Tip jar

  Tips for data

  by markw on Tue Jan 01, 2008 at 05:16:48 AM PDT

•  The DMR contradicts every other poll out there:

  Last 8 Iowa polls: (all released in the last week)

  CNN Opinion Research

  Clinton 33
  Obama 31
  Edwards 22

  Zogby

  Clinton 30
  Obama 26
  Edwards 25

  Des Moines Register

  Obama - 32%
  Clinton - 25%
  Edwards - 24%

  Insider Advantage

  Clinton - 30%
  Edwards - 29%
  Obama - 22%

  McClatchy/MSNBC

  Edwards - 24%
  Clinton - 23%
  Obama - 22%

  Reuters/C-SPAN/Zogby

  Clinton - 31%
  Obama - 27%
  Edwards - 24%

  Mason-Dixon

  Edwards - 24%
  Clinton - 23%
  Obama - 22%

  American Research Group

  Clinton - 31%
  Edwards - 24%
  Obama - 24%

  by phoenixdreamz on Tue Jan 01, 2008 at 05:20:07 AM PDT

  •  It all turns on the turnout model

    Weekend samples are notoriously difficult, too.

    by DCDemocrat on Tue Jan 01, 2008 at 05:29:51 AM PDT

    •  But the polls with Obama slipping...

      had weekend samples. Moreover, Obama's slippage tracks with the peak of the holiday travel season.

      I didn't here and Hillary people bitching about that.

      by sunbode2000 on Tue Jan 01, 2008 at 05:35:02 AM PDT

      •  So it would seem that we need to

        wait till Thursday.  By then, someone who was planning to go will have had a heart attack and won't go.  Someone else who wasn't planning to go will get a call from her best friend Sally and decide, "What the heck.  It could be fun."  We'll see what happens.

        by DCDemocrat on Tue Jan 01, 2008 at 05:43:05 AM PDT

        •  uh...yes. gotta wait.

          by sunbode2000 on Tue Jan 01, 2008 at 06:00:25 AM PDT

          •  i hate it, too

            by DCDemocrat on Tue Jan 01, 2008 at 06:02:52 AM PDT

•  This poll will help galvanize...

  the Obama folks.

  The most important thing is that this poll crushes the Edwards surge narrative.

  Indeed, this poll might be worth a couple points to Obama.

  by sunbode2000 on Tue Jan 01, 2008 at 05:20:42 AM PDT

  •  Ummmm...

    The most important thing is that this poll crushes the Edwards surge narrative.

    From the DMR's David Yepsen discussing his paper's poll:

    "the nightly results show Obama's support flat over the last two nights, Clinton's declining each night, and Edwards' support increasing each night."

    That's an interesting way to "crush" the Edwards surge narrative.

    by flubber on Tue Jan 01, 2008 at 05:31:14 AM PDT

    •  Uh yes....

      I think the spirits of Team Obama were getting down a bit---so yes, they needed this news to help galvanize their supporters/leaners.

      Nothing brings support like the belief that one can win.

      by sunbode2000 on Tue Jan 01, 2008 at 05:33:45 AM PDT

    •  If Yepsen says it

      It must be true.

      by ParaHammer on Tue Jan 01, 2008 at 05:34:49 AM PDT

      •  the fifth gospel

        by DCDemocrat on Tue Jan 01, 2008 at 05:43:33 AM PDT

    •  That's not what the poll itself said

      Obama goes up each night - on the last 2-day rolling average he was up to 34%, Edwards to 26%, and Clinton down to 23%.  It showed Clinton was declining and Obama surging, with Edwards tracking Obama.

      by clambake on Tue Jan 01, 2008 at 06:53:20 AM PDT

•  DMR used a different model

  which accounts for why their results are so different.

  Who's right?

  Well, we'll find out Thursday.

  by plf515 on Tue Jan 01, 2008 at 05:20:53 AM PDT

  •  Ok, but what's up with Iowa?

    I mean, why did they insist on mixing politics with the holidays and use a different poll model than everybody else?

    What are they thinking??

    by phoenixdreamz on Tue Jan 01, 2008 at 05:27:21 AM PDT

    •  What the hell was ARG doing?

      by sunbode2000 on Tue Jan 01, 2008 at 05:29:16 AM PDT

      •  Also a very good question yes

        by phoenixdreamz on Tue Jan 01, 2008 at 05:30:53 AM PDT

      •  ARG said

        that it's model assumed about 150,000 people show up to the polls.  I am not sure what difference that makes, but that was their model.

        by DCDemocrat on Tue Jan 01, 2008 at 05:41:15 AM PDT

        •  ARG MODEL

          Their model used a 134,000 turn out, the DMR TURNOUT MODEL IS 150,000.

          ny times indicated that all 3 campaigns agreed to the following turn outs:

          125000- 130,000 Edwards the winner
          130,00-140,000 Clinton the winner
          140,000 + Obama the winner.

          Finally turnout in 2004 was 125,000

          With the ground games of the 6 democratic candidates, Gov. Culver indicated that an increase of 20,000 or more is what he expects.

          by BDM on Tue Jan 01, 2008 at 06:34:25 AM PDT

•  Oh Sweet Polls Save Us

  from ourselves.

  by Detroit Mark on Tue Jan 01, 2008 at 05:24:40 AM PDT

  •  indeed

    by DCDemocrat on Tue Jan 01, 2008 at 05:30:04 AM PDT

•  the one who gets their voters to the caucus wins

  it's not that complicated

  by gaspare on Tue Jan 01, 2008 at 05:29:12 AM PDT

•  The two most accurate polls in IA

  are the Des Moines Register and ABC News according to Pollster.  Both are currently showing Obama leading.  However, ABC's poll is from mid-Dec.  I expect they will release a new poll prior to caucus.

  Obama's lead also reflects what is happening on the ground in Iowa.  Obama is drawing much larger crowds than Edwards and Clinton.  And Obama is still getting large amounts of undecideds to his events.  

  Lastly, I am hearing rumors of Biden and Kucinich making deals to have their supporters go to Obama if they are not viable.  Anyone else hear this?

  by askew on Tue Jan 01, 2008 at 05:48:43 AM PDT

  •  i saw a post

    on obama's website from someone claiming that kucinich was telling his folks to go to obama. now, i saw that a couple of days ago...

    but haven't seen/heard anything about that here in the media in iowa...don't know if it's true or not.

    by sunhaws on Tue Jan 01, 2008 at 06:39:48 AM PDT

•  Turnout will be 200,000+. Heres why-

  First of all, Hillary Clinton is running for president. Shes the most famous woman in the world. That in and of itself means lots of people coming out to vote. This isnt your typical milquetoast white guy running for president. Its Hillary FRICKIN CLINTON!!!

  Second, and also important, is the fact that Barack Obama is running for president. While this may not be significant as Hillary Clinton, the difference is that Barack Obama has been a turnout genious since 1992. He simply knows how to get apathetic voters to the polls and he's been doing it for a long time.

  The combined effects of this, are also added to the general drama and media highlight of a close campaign between three heavyweights. This motivates the party. Remember, 750,000 people voted for John Kerry in Iowa. So one shuoldn't be shocked by increased turnout for a highly contested caucus. The previous caucuses were so boring and the candidates so colorless and uninteresting to the general public that it did not nearly generate the interest that this caucus will.

  by brooklynbadboy on Tue Jan 01, 2008 at 05:58:23 AM PDT

  •  I don't doubt...

    I don't doubt that turnout will increase dramatically.  What I find strange is that turnout will increase amond Independents by 340%, whereas turnout will increase by Democrats by only 11%.

    Or, if you prefer to take a 300,000 turnout model (rather than 200,000), that has turnout increasing by independents by 509%, and turnout increasing by Democrats by 66%.

    Aren't Democrats at least as excited about a Democratic caucus as Independents are?

    by markw on Tue Jan 01, 2008 at 06:32:43 AM PDT

    •  I think so.

      The difference is that now we have a candidate who is actively going after independents. He did the same thing in Illinois, where he unexpectedly overwhelmed his competitors with big turnout from independents and republicans who came over to vote in the Democratic Primary in 2004. THAT is how you build the party, not by waiting until the general election. Many people do not realise Obama was not nearly the favorite to win that primary, but he did.

      Barack Obama isnt just campaigning to win 40% of the country. He's going for the big enchilada, and thats exactly what this party needs.

      by brooklynbadboy on Tue Jan 01, 2008 at 07:00:04 AM PDT

    •  the turnout was already high in 2004

      what folks who predict giant increases in 2008 turnout generally fail to recall is that turnout in 2004 was already up significantly from previous elections.

      it's not like you're starting from a low base.

      if turnout is only up 10% or so from 2004 it will equal a massive turnout in historical terms.

      to double 2004's turnout would be a completely unprecedented phenomenon.

      it is sort of unlikely, imo...

      by nbutter on Tue Jan 01, 2008 at 03:16:33 PM PDT

•  Thursday's weather: cold ... but clear.

  at least in Des Moines.

  highs in upper 20s, lows in low 20s.

  Sioux City looking a bit more cloudy, but a tad warmer.

  no snowstorm in sight, apparently.

  so...no one will be able to blame the weather.

  by GernBlanzten on Tue Jan 01, 2008 at 06:01:42 AM PDT

  •  They have been pretty

    consistent about predicting good (well, relatively speaking) weather for Thursday.

    by DCDemocrat on Tue Jan 01, 2008 at 06:10:41 AM PDT

    •  yeah, good as in "Iowa good" :)

      by GernBlanzten on Tue Jan 01, 2008 at 06:25:13 AM PDT

•  If those 2004 numbers are correct,

  it sure looks like the Republicans, already knowing who there candidate was going to be, crossed over to participate in the Democratic caucus and choose our nominee. The only fair way to pick a candidate is to have every states primary on the same day, in my opinion.

  by cackyp on Tue Jan 01, 2008 at 06:15:40 AM PDT

•  Hope Hillary finishes a distant third

  Well deserved

  by whitetiger on Tue Jan 01, 2008 at 06:15:44 AM PDT

  •  sorry, she's going to be second ; )

    JRE: 1

    HRC: 2

    BO: 3

    by nbutter on Tue Jan 01, 2008 at 03:17:17 PM PDT

•  Here's a unique analysis

  with the Caucus being so close to the holidays, it's impossible for any poll to predict the outcome.

  In the final analysis, one of the many divergent polls will be accurate through pure dumb luck and that poll will be touted as godlike in predicting Iowa come 2012.

  - Albert Einstein

  by Walt starr on Tue Jan 01, 2008 at 06:27:57 AM PDT

•  I don't doubt that...

  I don't doubt that turnout will increase dramatically.  What I find strange is that turnout will increase amond Independents by 340%, whereas turnout will increase by Democrats by only 11%.

  Or, if you prefer to take a 300,000 turnout model (rather than 200,000), that has turnout increasing by independents by 509%, and turnout increasing by Democrats by 66%.

  Aren't Democrats at least as excited about a Democratic caucus as Independents are?

  by markw on Tue Jan 01, 2008 at 06:31:02 AM PDT

•  It's all about the rural turnout

  and who they caucus for.  The Iowa Caucus is skewed to those areas, as they may have far fewer people.

  by funluvn1 on Tue Jan 01, 2008 at 06:32:43 AM PDT

  •  yep, and that's edwards country

    all three campaigns have been in rural counties but Edwards clearly is strongest there.

    by nbutter on Tue Jan 01, 2008 at 03:18:14 PM PDT

•  Pollster.com is the best place to look to predict

  the outcome.  Hillary's numbers have been fairly consistent in these polls, and pollster.com gives her a 4 point lead.  I think that is about right.  I seriously doubt that indies are going to increase their turnout for the Dems by that whopping margin. Therefore, I think it's Hillary #1, Edwards #2 and Obama #3.

  I think Obama has hurt himself among traditional democrats by pandering to republican and DC beltway 'centrist' frames on domestic issues, bipartisanship and 'all those things' as Arnold might say. Given that the people most likely to vote in a Democratic caucus are Democrats, then I think the polls which have a more natural distribution of voters among the three groups are probably more accurate.  

  by khyber900 on Tue Jan 01, 2008 at 06:46:21 AM PDT

•  The CNN poll actually validates the DMR poll

  It's true that Hillary is slightly ahead among Dems.  The CNN poll is just undersampling Obama's support.  There's every reason to believe turnout is going to be huge.

  http://www.chicagomag.com/...

  If Obama could register 150,000 new black voters in Chicago in 1992 with much less money, I'm sure getting a ton of new people to caucus is a piece of cake.

  by Fro on Tue Jan 01, 2008 at 06:46:36 AM PDT

  •  150,000 new voters in 1992

    I never realized the number was that high.  

    by askew on Tue Jan 01, 2008 at 11:07:29 AM PDT

•  Agree

  I think the turnout model is too high.

  by Paleo on Tue Jan 01, 2008 at 07:04:02 AM PDT

•  I'll say this Edwards is not having a good

  morning the weather looks good for caucus night, the DMR poll which is the only one Iowan's pay attention to will really help Obama create momentum
  and although Clinton does better in some other polls and turnout models Edwards does not. the talk of Edwards suring is now over. Anti-Clinton voters  who are undecided seeing those numbers will probably break to Obama.

  by nevadadem on Tue Jan 01, 2008 at 07:14:23 AM PDT

•  Nothing going on

  DMR is the gold standard of polls in Iowa
  CNN is the Clinton News network and obviously their polls cant be trusted

  by whitetiger on Tue Jan 01, 2008 at 07:54:55 AM PDT

•  Interesting caucus attendance possibility:

  In my town in Maine in 2004, I had the impression that attendance for the Dem caucus had doubled over the 2000 number to about 7% of registered voters (in seriously crappy weather, IIRC).  It also seemed like the caucus "veterans" overwhelmingly supported Kerry because he was "most likely to win the General."  Whereas most of the "newbies" (mostly first-timers) supported Dean or Kucinich.

  So, in Iowa in 2004,

  Total caucus attendance was about 6.5% of registered voters.
  Dem participation was about 5% of RV.
  IND participation was about 1% of RV.

  Projections are for an increase to
  10% of RV for total caucus participation
  5.5% of RV for Dems
  4% of RV for IND (adding another 2%-3% to Dem total?)

  But what if the weather's decent and the Dem total actually doubles?  How might this affect the outcome for the "non-establishment" candidates (Obama, Edwards, Kucinich)?

  I also often wonder how pollsters project "likely caucus goers".  How much do they adjust "likelihood" by whether or not some one has previously participated in a caucus?

  by sxwarren on Tue Jan 01, 2008 at 08:19:26 AM PDT

•  can you lose dems and win the caucus?

  if Obama actually wins the Iowa caucuses while losing among iowa democrats it will be a hell of a story!

  by nbutter on Tue Jan 01, 2008 at 03:19:43 PM PDT