Pollster.com

Articles and Analysis

 

POLL: ARG Iowa Caucus


A new American Research Group statewide survey of likely caucus goers in Iowa (conducted 12/20 through 12/23) finds:

  • Among 600 likely Democratic caucus goers, Sen. Hillary Clinton leads former Sen. John Edwards (34% to 20%) in a statewide survey; former Sen. Barack Obama trails at 19%, Sen. Joe Biden at 8%, Gov. Bill Richardson at 5%.
  • Among 600 likely Republican caucus goers, former Gov. Mike Huckabee runs at 23%, former Gov. Mitt Romney 21%, and Sen. John McCain 17% in a statewide primary; former Mayor Rudy Giuliani trails at 14%, Rep. Ron Paul at 10%.
  • All other candidates receive less than five percent each. The margin of error is four percent for both subgroups.

 

Comments

Intrade has Obama plummetting......

____________________

Tommy Ates, Austin, TX:

The only people who pay attention to ARG is Hillary Clinton herself. ARG has never in Iowa by less than 9%, nationally nearly always double the other polls. ARG has been discredited over the owner's political ties to the Clinton campaign. Oop! He is a main contributor. Could that be why every ARG poll is an 'outlier'?

____________________

Anonymous:

American Research Group is a respected national polling firm and their numbers are normally fairly accurate.

____________________

Mervin:

Why are you hating and see the polls for what they are......It really astonishes me that no one in the media or even haters out there,they dont have a problem when obama is leading to make it a hot story..Now we have this poll,everyone wants to bury themselves and cry foul....But thats exactly whats gonna happen in Iowa,Hillary will bury Iowa......Happy Holidays.....

____________________

barnstormer1:

When the polls favor Obama, they are valid. When they favor anyone else, watch out --they are rigged. Obama's rise is a creation of the media.

____________________

terrondt:

the haters don't like the result of the poll so they cry like babies.

____________________

PJ from NYC:

Hahahahaha...
these polls mean nothing...
hahaha...they called me too...and I am from NYC
sorry for everyone...
lets go Ron Paul

____________________

Stephen Hendricks:

Hasn't anyone else noticed that ARG reported a separate poll from 12/16 through 12/19 with some radically different results?

Does anyone really believe that Clinton gained five percent and Obama lost six percent almost literally overnight? Does it seem credible that Ron Paul's support would zoom from 4% to 10% in a matter of 72 hours? And did Huckabee really lose five points while Romney gained four points in the same period?

The results suggest not just one outlier result, but two.

____________________

Andrew:

ARG has been good in the past, according to the research I've done. Not superb, but good. Claims that ARG is this terrible wacky polling company are exaggerated.

____________________

Paul:

Let's look at the numbers. Obama's December 2007 Iowa numbers from polls which I track have been 33 (R2K), 27 (Diageo), 25 (MSNBC), 27 (RAS), 26 (RAS), 35 (Newsweek), 25 (ARG), 24 (ZOG), 30-33-32 (IA), 28 (CNN), and 26 (Pew). This ARG number, 19, is clearly far outside the range we have seen recently. Obama seemed to have settled in the high 20's or low 30's, but nothing close to a high teen.

I would also look closely at the McCain number in this ARG poll (17). In the last five days, McCain's number ranged from 6 to 9.

When I see one number (Obama) which looks "off the charts", I would say just a statistical anomaly. But when I see a 17 for McCain, I say the samples have to be significantly unlike samples in comparable polls. How else do you explain 10+ point differences?

____________________

Tommy Ates, Austin, TX:

Paul,

They can't. I have never trusted ARG for anything. Never. And they still haven't addressed the fact that ARG's owner is a top contributor to the Hillary Clinton for President campaign.

____________________

Andrew:

Paul, what evidence do you have that ARG's owner is a contributor to Hillary Clinton's campaign?
If true, this automatically disqualify this poll from being trustworthy.

____________________

Tommy Ates, Austin, TX:

Andrew,

Here is one of many links:

CNN pollster Vinod Gupta to host Hillary Clinton fundraiser
FreeRepublic.com | November 25, 2007 | HAL9000

Posted on 11/25/2007 3:13:18 PM PST by HAL9000

Last January, an exclusive report was published on FreeRepublic.com, revealing CNN had hired a major Hillary Clinton backer named Vinod Gupta to conduct their presidential polling for the 2008 election.

Since then, several reports have appeared in the mainstream media, and on this site, detailing the antics of Gupta with the Clintons - the millions of dollars in corporate funds and services he has given to them, his conflict of interests in acting as CNN's pollster after declaring that "she'll be our next president", etc.

As the negative publicity about him grew, Vinod Gupta eventually claimed that he had severed his ties to Hillary Clinton's campaign -

http://www.freerepublic.com/focus/f-news/1930370/posts

____________________

Diggs:

Maybe this outlier is at least partially a result of holiday travelers. We know that college students are on break now and young people that aren't college students are more likely to be traveling during the holidays, which means a significant portion of Obama's supporters might be out of town. A possible explanation, or at least in part.

On a related note, do these polling outfits adjust their numbers at all to calculate for cell phone users that aren't polled, or is that still an issue?

____________________

Mark Blumenthal:

The American Research Group is run (and presumably owned) by Dick Bennett. I find no record of any contributions by Bennett to the Clinton or any other presidential campaign on opensecrets.org. Searching online, I also find no connection whatsoever between ARG and either InfoUSA or Vinod Gupta.

Bennett also responded here to the question "who is behind ARG?"

____________________

BDM:

I cannot find any ARG polls for the IA caucus OF 2004.

wHAT HAS BEEN THEIR RECORD FOR POLLING THE IA CAUCUSES? DOES ANYBODY KNOW?

____________________

dcshungu:
The American Research Group is run (and presumably owned) by Dick Bennett. I find no record of any contributions by Bennett to the Clinton or any other presidential campaign on opensecrets.org. Searching online, I also find no connection whatsoever between ARG and either InfoUSA or Vinod Gupta.

Bennett also responded here to the question "who is behind ARG?"

Posted by: Mark Blumenthal | December 25, 2007 9:11 AM


Mark,

Thank you for doing the leg work so that we would not have to. Pollster.com has been my to-go-to site for polling news and in-depth analyses, which means that you were already on-board as a site to trust, but your getting up on Xmas day to set the record straight raises you another significant notch in my esteem because it is so above and beyond the call of duty...


Merry Xmas!

____________________

Mike S.:

Out of all the polling firms shown on pollster.com, ARG appeared to consistently show the lowest numbers for Ron Paul (except for Survey USA who still completely excludes him, which baffles me). Yet, in this poll, he's up to 10 percent. I know he's not the frontrunner, but this jump should not be marginalized. He might shock some people with how well he places in Iowa and NH.

____________________

wxdavid:

The ARG poll is crap NOT because its founder may or may not have connections to Hillary. Its crap because of TWO very good reasons

1) ARG polls results are WAAAAY off in Iowa. ANY poll...regardles of whether its Strategic
Vision or Galup or ABC Ipsos that has Ron Paul at 10%... or shows HUGE overnight gains for Hillary and a drop of Obama...

COULD BE RIGHT.

But the ARG poll has NO other support. IF 7 recent polls have Obama betwen 25% and 32% and One poll (regardless of which one) has him at 19%... why would ANYONE care what that poll said?

and what about McCain Iowa iowa getting a very strong 17% with NO campaigning there? if that is correct that is BIG new for McCain

2) ARG track record in in 2004 primaries and General election was pretty poor.

____________________

wxdavid:

The ARG poll is crap NOT because its founder may or may not have connections to Hillary. Its crap because of TWO very good reasons

1) ARG polls results are WAAAAY off in Iowa. ANY poll...regardles of whether its Strategic
Vision or Galup or ABC Ipsos that has Ron Paul at 10%... or shows HUGE overnight gains for Hillary and a drop of Obama...

COULD BE RIGHT.

But the ARG poll has NO other support. IF 7 recent polls have Obama between 25% and 32% and One poll (regardless of which one) has him at 19%... why would ANYONE care what that poll said?

and what about McCain Iowa iowa getting a very strong 17% with NO campaigning there? if that is correct that is BIG new for McCain

2) ARG track record in in 2004 primaries and General election was pretty poor.

____________________

wxdavid:

The ARG poll is crap NOT because its founder may or may not have connections to Hillary. Its crap because of TWO very good reasons

1) ARG polls results are WAAAAY off in Iowa. ANY poll...regardles of whether its Strategic
Vision or Galup or ABC Ipsos that has Ron Paul at 10%... or shows HUGE overnight gains for Hillary and a drop of Obama...

COULD BE RIGHT.

But the ARG poll has NO other support. IF 7 recent polls have Obama between 25% and 32% and One poll (regardless of which one) has him at 19%... why would ANYONE care what that poll said?

and what about McCain Iowa iowa getting a very strong 17% with NO campaigning there? if that is correct that is BIG new for McCain

2) ARG track record in in 2004 primaries and General election was pretty poor.

____________________

Lupe:

You guys are missing the point.

The reason this poll is crap is because it's impossible to poll this close to Christmas without biasing your sample. It's not just that literally something like 40-50% of the country is displaced from their usual place of residence. It's that not all people are equally likely to travel. For example, virtually every student travels for Christmas. On the other hand, senior citizens are only about half as likely to travel as younger Americans. And this is common sense; you go to your grandma's house for Christmas, rather than the other way around. Since older voters tend to be Clinton's best group whereas Obama does best among younger voters, you'd expect any polling taking place during this period to be favorable to her.

That's why none of the "name brand" polling outfits are releasing polls right now.

Would it be possible for a polling outfit to make adjustments to these numbers to correct for these effects? Perhaps. But it's unlikely that they'd correct for them entirely, and more to the point, it isn't clear that ARG, which has released no cross-tabulations nor any information about their methodology, as tried to correct for them at all.


____________________

Andrew:

wxdavid, Mr. Blumenthal proved that there is absolutely not even a hint that the ARG management is linked to Hillary Clinton. Some Freeper who was apparently not sober came here talking about a person from Opinion Research, who is another totally different company. So don't come here say that ARG "may or may not" have links to Hillary. It doesn't.

And no, ARG's record is not very poor.
And no, a poll's accuracy is not measured against other polls without taking election results into consideration.

____________________

Tommy Ates, Austin, TX:

March 31, 2005
Disclosing Party ID: American Research Group

Here is another response to my query to pollsters who do not typically report party identification in online releases to explain that policy. Today we hear from Dick Bennett, president of the American Research Group (ARG):

We have provided party registration or party ID for almost all of our political surveys posted online and because of your post, we will include it for all our political surveys posted online.

Our interviewing system has a look-up function by area code which places a party registration question ("Are you currently registered to vote as ...") in the screen (at the beginning of the survey) for states with party registration and a party ID question ("Do you consider yourself to be ...") in the screen for the other states. Party registration and party ID from this process get combined in national surveys.

We are currently asking a political party point-of-view question near the end of surveys and the responses to that question show much more survey-to-survey movement than the registration question (which is very stable) and the ID question (less stable). I can't tell you if it is the question or question order.

[Emphasis added]

Thank you, ARG.

It is important to note that what ARG does, as explained above, is very different from the way other pollsters ask about party ID. ARG asks about party registration on some states, party identification in others and then combines the two results into a single variable. Whatever the merits of this approach, the will results not be comparable to those of other polling organizations.

http://www.mysterypollster.com/main/2005/03/disclosing_part_4.html

____________________

BSM:

ARG did not provide polling for the Iowa caucuses in 2004.

This is the first cycle that they have polled for the IA caucuses, thus they have no prior record for polling Iowa's presidential caucuses.

____________________

Angry Commenter:

Why are the grammar so bad in this comments?

____________________

Jose Padilla:

A common complaint seems to be that ARG polled over the Christmas holidays when a lot of students are gone. But aren't the caucuses on Jan. 3? Won't most of those students still be gone then?

____________________

Paul Hollings:

Interesting that ARG's New Hampshire poll has Hillary up by 14, whereas three other contemporaneous polls have it essentially tied (Globe has Obama +2, Gallup has it tied, and Rasmussen has it Clinton +3).

Either ARG is onto something the others aren't, or they're going to have mud on their faces.

____________________

Paul Hollings:

In NH, ARG has Clinton +14, whereas three other contemporaneous polls have it very close (Globe has Obama +2, Rasmussen has Hillary +3, and Gallup has it tied.)

In SC, ARG's late November poll has it Clinton +24. All other polls around that time and since then have it very close.

Makes you wonder about ARG - are they trying to influence the outcome? Do they know something that no other pollsters know? Are they just incredibly incompetent?

____________________

Paul I.:

I'm probably the only regular visit to this great site who *doesn't* have the primary dates memorized...if those dates could be added somewhere to the polling graphics it would be much appreciated!

____________________

Anonymous:

There are reasons for this latest surge, which are commercials in heavy rotation touting the Des Moines Register endorsement of Hillary and another one narrated by Bill Clinton. Don't underestimate Bill Clinton's popularity or the influence of the Des Moines Register.

____________________

Nate:

http://www.washingtonpost.com/ac2/wp-dyn/A41186-2004Jan23?language=printer


ARG had McCain losing to bush in NH - only to see him win by almost 20 points.

Their methodology is suspect to say the least - I have no idea bout Bennett's political leanings, but that is meaningless. There is no way that there are 15 point swings that often in Iowa as his polls have suggested.

http://ajacksonian.blogspot.com/2007/12/problem-with-american-research-group.html

____________________

GordonsGirl:

Maybe this should be considered for balance:
/08-IA-Dem-Pres-Primary.html

____________________

Robert:

It's possible the difference is accounted for by the students who didn't answer the phone because they were at home for Christmas. Hopefully for Obama, they'll be back by the 3rd.

____________________

It's possible the difference is accounted for by the students who didn't answer the phone because they were at home for Christmas. Hopefully for Obama, they'll be back by the 3rd.

____________________

Jacob:

That last comment should be taken down immediately. It's complete BS and definitely not a substantive addition to this discussion.

____________________

dcshungu:

Angry Commenter asked on December 26, 2007 10:37 AM


Why are the grammar so bad in this comments?

LOL.Maybe because most people are posting "on the fly", with minimal proofing/error checking?...

BTW, since you asked, would you care to tell us what your excuse is?... LOL.

____________________

Jo Ann:

Why is anyone even paying attention at all to any poll? Who cares? I care more about coverage of the candidates and their stands on the issues which have been TOTALLY lacking this election cycle! Let's forget the horserace coverage and actually get some real news!

____________________

Paul:

In response to the question about primary dates, best source I believe accurate is the following:
http://www.vote-smart.org/election_president_state_primary_dates.php

____________________

George:

Read more on ARG and Mr. Bennett at this site, and form your own opinions about this "Research"
company.

http://ajacksonian.blogspot.com/2007/12/problem-with-american-research-group.html

Can anyone tell me what is meant that their poll data is within +/- 3%, 95% of the time. Is this even statistically significant?

____________________

edgeways:

heh, looks like the "haters' where correct in this instance. ARG seems a little batty. I ran across it looking for poll stats for the WI primary and they are the only ones predicting a Clinton lead, and a large one at that, while all the other sites have Obama ahead by between 4 and 12 points. I wonder at ARG's methodology and how the results are so different from other places? Indeed, after tomorrow I may file this place down as being unreliable and always ignore it going forward.

____________________

Alan Z.:

I've read the blog that George posted above because I was likewise suspicious of the Wisconsin poll results by ARG. The fact that the ARG site has no company information is a clear indicator that there is something fishy. Any reputable organization would include detailed information about the organization and their management. Good example of how badly they can be off, was South Carolina - by almost 20 points just a couple of days before the primary.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR