Two Different Polls, Two Completely Different Results

By: Lowell
Published On: 12/19/2007 7:18:59 AM

I agree with Jerome Armstrong that "Polling in Iowa is Bizzare."

Two polls come out of Iowa.

The one blogged last night here, from Insider Advantage shows Edwards (30), Clinton (26), Obama (24).

The poll done by TNS and released at the same time shows Obama (33), Clinton (29), Edwards (20).

desmoinesdem is right about polling in Iowa being very difficult, but wtf, one of these polls is drastically wrong.

That's right, either John Edwards is in first place in Iowa or he's in last. Either Barack Obama's in first or he's in last.  In fact, the only result these two polls agree on is that Hillary Clinton is in second place in Iowa.  

I'm really starting to think that none of these Iowa polls are particularly helpful, particularly given the extreme difficulty of figuring out which 100,000 or so voters will actually turn up in the Democratic caucuses on January 3.  Most importantly, perhaps, is the turnout -- or lack thereof -- of young voters, where Barack Obama is particularly strong, and seniors, where Clinton and Edwards are strong (and Obama is weak).  In other words, who the heck knows?!?

P.S. Will the key to this election be the fact that seniors will almost certainly vote, while college students and other young people are not as certain?


Comments



momentum ... (bcat - 12/19/2007 8:43:16 AM)
It's all about timing and momentum. In other words: any one poll won't be terribly informative. You have to look at polling trends, and those trends show Obama rising in a pretty dramatic way:

http://www.pollster.com/08-IA-...
http://www.realclearpolitics.c...

Of course Iowa is notoriously difficult to predict, because the caucus process doesn't exactly lend itself to obvious conclusions. For example: if Richardson and Biden fail to hit 15% on January 3rd, will their supporters go to Edwards, Clinton, or Obama? That could throw the race to anyone, even a distant third like Edwards.



Union strong (Terry - 12/19/2007 9:08:30 AM)
Edwards is very strong among union members who historically turn out in the caucuses. He is also strong in the rural areas that Clinton has neglected in favor of the bigger urban areas like Des Moines. And, Edwards is the second choice for the majority of voters who are currently for a second tier candidate..I'm with Newsweek. Edwards is the sleeper candidate that has been pretty much ignored by the media and has a really good chance of coming in first place. If Obama finishes a close second and Hillary is third, NH's independents might be the key vote for momentum going forward. My cousin lives there now and is an Independent and is still undecided. She said they only thing she knows for sure is that she is sick to death of the mail and phone calls.  


You're largely right; they're worthless. (Silence Dogood - 12/19/2007 11:01:45 AM)
At the root of this problem is what the pollster in question classifies as a likely primary voter/caucus goer.  Especially early on, the polls that showed Obama trending upwards had a much more generous opinion of who was a likely caucus goer, although I'm inclined to think that he is probably is making significant headway with voters who, like, actually are going to go to the caucus and aren't 26 year olds who say they're going to the caucus but haven't voted in the last three federal elections?

I wonder who the heck TNS thinks is a likely caucus goer?  I also wonder if they fudged their poll numbers so news agencies would write about it because the media loves conflict and controversy, and TNS loves potential clients seeing its name in the papers.  There was a polling company about five to six months ago that was projecting huge leads for Obama over Hillary in New Hampshire that conflicted with about four other polls, and the reason is they intentionally fudged their definition of likely primary voters to reflect an outlier for people to report on.



Edwards Edwards Edwards (citizenindy - 12/19/2007 11:08:14 AM)
From the Washington Post

Adding to the challenge for Clinton and Obama is that they are relying more heavily than Edwards on potential first-time caucus participants. More than half the supporters of Clinton and Obama have never caucused, while two-thirds of Edwards's backers have done so. Edwards is hoping to draw on the network he built in Iowa four years ago when he finished second to Sen. John F. Kerry, the eventual Democratic nominee.

If that is true I'm almost ready to call it for Edwards



How Super Tuesday could NOT decide, and VA stays relevant.. (cdgoin - 12/19/2007 1:37:35 PM)
I have to agree.. I think Edwards will win.. Obama second, and a virtual tie (Say less than a 4% spread) between Richardson and Clinton.

I would call it this way:

Edwards - 31
Obama - 23
Clinton - 21
Richardson - 19
Biden - 4
Kucinich - 1
Uncommittted/Other - 1

Now based on that conjecture here is a way this whole thing goes into a free for all and proves once again the media is full of you know what.. and that it could NOT be over super tuesday... !

>>>

With the loss by both Clinton and Obama - The news will focus on how well Edwards and Richardson did.. and how there really is large portion of the party that doesn't want Hillary.

This will throw New Hampshire into a polling nightmare. The media will be all over the place and so will the polls.

New Hampshire call:

Clinton - 29
Edwards - 27
Richardson - 21
Obama - 19
Others - 2

The media doesnt know what to do with itself.. report on a "comeback" by Clinton or the ascent of Edwards and Richardson ? In any case they stop trying to annoint a candidate and start saying anyone could win.  

With this, the media finally gives Governor Richardson prime time and more than the casual mention at the end of a story. Due ot this his numbers jump in the polls as people realize he is the only candidate with real experience.

Michigan throws another log on the fire.. (although one that doesnt technically matter but it gives the media two days of news to report)

Uncommitted - 52
Hillary Clinton - 26
Dennis Kucinich - 12
Other/Write-in - 10

Since these will be the only candidates in the race and on the Michigan ballot.. the fact HIllary loses by more than 50% to "Uncommitted" throws the media into a new frenzy..

As Clinton and Obama rush to South Carolina (Now being called the firewall for both) Gov. Richardson focuses on Nevada and Nevada gives him a huge upset..

Richardson - 31
Edwards - 25
Obama - 24
Clinton - 19
Others - 1

The Obama and Clinton machines are in full force in South Carolina..

Obama gets his first win.. - 29
Edwards - 25
Clinton - 23
Richardson - 21
Others - 2

But it clearifys nothing.. other than a free fall for the Clinton campaign..

Florida..

Florida like Michagan doesnt mean anything..

BUT everyone is on the ballot.. thus its newsworthy..
Especailly since Clinton "bounces back"

Clinton - 31
Richardson - 26
Obama - 22
Edwards - 17
Others - 4

Proving its STILL a 4 way race..

Edwards (IA - 57)
Clinton (NH - 30) & (FL*)
Richardson (NV - 33)
Obama (SC - 54)
Uncommitted (MI*)

SUPER TUESDAY:

Clinton - 654 Delegates
Arkansas 47
Connecticut 60
Delaware 23
Idaho 23
Massachusetts 121
Minnesota 88
Democrats Abroad 11
New York 281

Richardson - 643 Delegates
Arizona 67
California 441
New Mexico 38
North Dakota 21
Oklahoma 47
Utah 29

Obama - 457 Delegates
Alabama 60
Illinois 185
New Jersey 127
Tennessee 85

Edwards - 321 Delegates
Alaska 18
Colorado 71
Georgia 103
Kansas 41
Missouri 88

Something like this happens and one of the four drop out.. and it becomes a state by state fight.. where Virginia is important afterall..

Could this happen? Possible not Probable..

If it does how it ends is anybodys guess..

But I rather have this than a media coranated Clinton... if she wins so be it.. but I want to get through Super Tuesday without her having even gotten close to winning it all..

Even more interesting is if you go through all this and lets it goes back and forth come March 2nd.. if Richardson is still in there is Texas (228) and for Obama or Clinton Ohio (161) that could make the difference.

Charles



Richardson (citizenindy - 12/19/2007 2:38:56 PM)
What is your source for him doing so well?

More on Edwards

"8a. (IF NAMED RICHARDSON, BIDEN, DODD, OR KUCINICH) If the only three viable candidates in your precinct were (Clinton), (Edwards) and (Obama), who would you support as your second choice?

Clinton Obama Edwards
16 29 41"



Ugh (cdgoin - 12/19/2007 6:23:10 PM)
Don't you hate when you close a window by mistake after writing something..

The simple answer is I have seen some county by county polling.. Also data based on caucus size, location, and delegates. I think Richardson will do much better in the rural areas than many expect.

He has the same things that Edwards have in his favor. strong support with his core voters, and both have a large percentage of supporters that have caucused before.

With Richardson holding a steady 12%... it shouldn't take much in rural areas for him to get 15% and be viable in many smaller caucuses. Just like Edwards in 2004. Also it doesn't take but one or two people to change 12% to 15%.

As for second choice.. it's obvious that Edwards, Richardson and Obama will gain from second choice balloting. Richardson maybe more.. if Edwards and Obama are already viable in a certain caucuses, a lot of the time caucus goers will go with the candidate that is close to viability to help them over the front runners. Hillary will be viable in most all caucuses statewide, so she will not benefit from second choice much at all.

Also there is strong "anyone but Hillary" movement in Iowa. Similar to the Anti-Dean movement that started about the same time last cycle. So, I dont think Hillary is near as strong as some would think. I think she will follow the same pattern of voting Dean did... same holds true for Obama. The Two groups that "supported" Dean in the polls prior to the Iowa collapse support them both. The young voters, college students, and those that support whomever the media says is winning. All unreliable. Then there is the negative campaigning and throngs of imported supportes to overwhelm the poor voters of Iowa.. they like to rebel against that. I see Richardson as getting the anyone but those three votes..

WHEN the polls ask whom of the top four or whom of ALL they support. In most the "second choice" polls, often Richardson is in the 14-20% range. Case in point the poll you mentioned only asked for an answer for the three viable candidates. When they add Richardson he comes in a close third to Obama, at 20-25% with Hillary under 15%.

So, lets say Richardson gets 20-25% of the delegates from the Rural areas.. and say maybe 10% in the metro areas.. he should be about 19% by the end of the day.. Edwards will do similarly well statewide and pick up most of the Richardson votes in caucuses that are not viable.

So I think he will pick up 4-5% over his polling numbers. or about 31%

Give Biden (very few peoples second choice) a few points (4%) and Kucinich the same as last time (1%)

So once you do the math.. that leaves Hillary and Obama fighting over the same group in metro areas and I think Obama will win that battle..

So viola.. my numbers..  

Edwards - 31
Obama - 23
Clinton - 21
Richardson - 19
Biden - 4
Kucinich - 1
Uncommittted/Other - 1

If this heppens.. you can bet the media will go nuts trying to explain it all.. especially how Richardson could be so close to Clinton and Obama, and how Edwards could win by such a margin..

That could easily make the results of New Hampshire very fluid as well..  



I hope Edwards pulls it off. (thegools - 12/20/2007 1:04:13 AM)
I am not a big fan of the idea of Hillary winning.  Obama would be fine though.


One big flaw in the polls (True Blue - 12/19/2007 11:54:02 AM)
They aren't showing undecideds.

30%+ of Democrats are undecided and 45%+ of Republicans.

What do these levels of undecided mean with only a couple of weeks left before Iowa?



Wait just a second... (Kryndis - 12/19/2007 12:03:38 PM)
They're not using butterfly ballots for these polls, are they?


Hey! (Lowell - 12/19/2007 12:45:18 PM)
That might explain it! :)


Lets not forget (cdgoin - 12/19/2007 12:35:12 PM)
If you think the polls in Iowa are hard to quantify..

Nevada is also a caucus state and they dont have much of a history.. so I would think the polls in Nevada are completely wrong.

Not only that, its a busy time of year for Vegas.. so who knows who will even go to the caucus.

Charles



It's simple (Captain Burke - 12/19/2007 3:57:55 PM)
If it is a listed sample poll calling only people who have participated in a caucus or low turnout election, it will skew older and likely be more accurate.  If it is a random digit sample where the firm does a lot of interviews (TNS and Newsweek) and ask likelihood to vote, it will include a lot of Independents and younger voters and not quite match who will actually vote.  Edwards does better with listed sample surveys and poorly with RDD ones.

This is one case where averaging does not help and is more garbage in garbage out.  Pollster.com does a good job in trying to collect the demographics that matter and age in all of these polls are all over the map.



NH (phillip123 - 12/19/2007 5:28:02 PM)
There is no way in Hell Richardson finishes ahead of Obama in NH.


I didn't say it was probable only possible (cdgoin - 12/19/2007 5:51:01 PM)
The dynamics of the race can change overnight from Iowa to New Hampshire. Depending on what happens in Iowa and how the media plays it..

Lots of uncommitted, independant, and undecided.. if Richardson comes out of Iowa as a viable candidate and Edwards wins.. Who knows how that will affect the voting in New Hampshire and Nevada. Especailly if Hillary comes in third in a close tie with Richardson.

All I know is I hope its a lot more interesting than watching a Clinton / Obama duel till Feburary. If thats all we get out of Iowa, NH, Nevada and SC..  then the media won and I give up on our "Democracy"..

Its almost gotten to the point with the media so controlled by special interests and they just tell everyone who to vote for and they do it..

At least in Russia they dont hide it.. here we try to act like we have a "Democracy"... but it's about as democratic now as Russia.



Updated Iowa Rasmussen (PM - 12/19/2007 8:50:14 PM)

Democratic Iowa Caucus
Iowa: Clinton 31% Obama 27% Edwards 22%
Wednesday, December 19, 2007

Those who want polls to tell them who will in Iowa will be sorely disappointed. Right now, the race is too close to call and the difficulties of polling during the upcoming holiday celebrations will not clarify the matter.

The latest Rasmussen Reports poll of the Iowa Democratic Caucus finds that Hillary Clinton is supported by 31% of Likely Caucus Participants. Barack Obama earns the vote from 27% and John Edwards is the top choice for 22%.

However, among those who are "certain" they will participate in the caucus, it gets a bit closer-Clinton 29% Obama 28% and Edwards 22%.

Among those who have participated in the caucus before it's Obama 26% Clinton 25% and Edwards 24%.

***
Adding to the closeness of the race are the second choice rankings. Edwards is the second choice for 28% while Obama is the number two pick for 22%. Clinton and Bill Richardson are the second choice for 15%. Second choices are important for two reasons. First, because 26% of Likely Caucus Participants say they might change their mind. This includes 8% who say there's a good chance they could change their mind.***

These results are generally similar to those from a week ago and the overall message is the same-it is still quite possible for any of the three leading candidates to emerge victorious.


http://www.rasmussenreports.co...