Do National Polls Really, Really Inflate Clinton's Lead? Really?

Back on Thursday morning, I posted an article claiming that most national primary polls inflate Hillary Clinton's lead. My argument had two basic points. First, national primary polls are sampling all voters who self-identify or lean Democratic, which is about 40-50% of the registered voting population, rather than the 15-20% of registered voters who actually comprise the Democratic primary electorate. Second, I argued that the 15-20% of registered voters who actually comprise the Democratic primary electorate are significantly less pro-Clinton than the 40-50% of registered voters who comprise all Democratic self-identifiers and leaners. I based my second point on evidence showing Obama performing stronger among high-income, high education, and high news consumption voters, as well on polls showing both Edwards and Obama performing better among "likely" primary / caucus voters than among "registered" Democrats and Democratic leaners.

Mystery Pollster, while agreeing with the first point,that national primary polls are sampling far too broad a spectrum of Democrats, takes issue with my second point, that this inflates Clinton's advantage, which he ultimately thinks lacks conclusive evidence. He writes:
I am not ready, however, to agree that this practice inflates Clinton's lead in national polling. What evidence I see is sketchy and contradictory. For example, Tom Riehle of RT Strategies recently shared some tabulations combining data from two recent surveys conducted by his company (in February and late March) for the Cook Political Report. Clinton wins 41% of the vote against Obama, Edwards and the other candidates on these two surveys. However, she wins more support from pure Democratic identifiers (44%) than among the independents that lean Democratic (33%). That difference is statistically significant despite the small sample sizes (n=558 Democrats, n=164 Democratic leaners). So if sampling too many voters means too many independents, it will tend to depress Clinton's vote rather than exaggerate it.

On the other hand, we did a quick comparison of national surveys fielded this year that asked the primary question of all adult Democrats (and leaners) versus those that asked the question of only registered voters that identify or lean Democratic. We looked only vote questions testing the whole field, but excluding Al Gore. Clinton received an average of 44% on four surveys conducted by Gallup and ABC/Washington Post that included all adult Democrats, and 38% on eleven surveys** that included only Democrats registered to vote (or "likely" to vote in the general election). So we have sketchy support for the notion that surveys with slightly tighter screens show Clinton with slightly less support.

Admittedly, there are many potential pitfalls with such comparisons (commenters, have at it), and none of the available data allows for a direct test of Bowers' contention. In other words, we have no survey that allows us to compare the vote among all Democratic identifiers to a theoretically "true" likely primary electorate. So I think the jury is still out.

Chris cites some data from recent LA Times/Bloomberg [PDF] and Pew Research Center surveys showing Obama running much closer among well educated Democrats. And better educated adults tend to vote at higher rates than less well educated voters. That much is true, although education is just one predictor of turnout. Another is age, and the Pew survey shows Clinton with more support among older voters who also tend to turn out at higher rates.
Fair enough. I even looked at the Rasmussen reports polls I cited as evidence for my argument, and it turns out their primary polling universe is drawn from all Democratic self-identifiers (774 likely voters) taken from a four day sample (2,000 likely voters) of their presidential approval tracking poll. In other words, even Rasmussen is drawing from 38-39% of the entire likely voter population for their primary voter sample. This might be too high, it might not be. It all depends on how Rasmussen determines "likely voters." Since, I do not know how Rasmussen determines likely voters, and since polling companies are loathe to give out information on their likely voter screens, doubt is even cast on that element of my evidence.

However, one aspect of my argument cannot be doubted. By sampling between 40-50% of all registered voters, the vast majority of national Democratic primary polls are not specifically sampling the Democratic primary electorate. As such, these polls should not only be taken with a grain of salt, but should almost be dismissed entirely as useful indicators of the current state of play in the Democratic primary / caucus season. If they are not polling the Democratic primary / caucus electorate, or even coming close to doing so, then they should not be used as indicators of opinion in the Democratic primary / caucus electorate. To use an analogous example, one does not poll the entire nation as a means of determining public opinion in California. One does not poll all Democratic self-identifiers and leaners in order to determine the opinion of Democratic primary and caucus goers.

News organizations commissioning national primary polls should either pony up the cash necessary to accurately sample the national primary electorate, or they should stop pimping their own polls as accurate reflections of the national primary and caucus electorate. Further, the American Association for Public Opinion Research should really lay into new organizations that don't do this, since it is blatant false advertising and manipulation of data by the national media. Finally, there may not be enough evidence to conclusively demonstrate my thesis at this time, but I still believe that when national polls sample an accurately narrow universe of the Democratic primary and caucus electorate, Clinton's lead will come in at 12% or less in every single survey. I just hope some organization will begin conducting the necessary (and expensive) surveys necessary to either prove or disprove my claim.

Tags: Hillary Clinton, inflated Clinton poll theory, polls, President 2008, Primary Elections (all tags)

Comments

23 Comments

well
what about the theory that the more the state is engaged in the campaign the worse that both Rudy and Hillary do? look at the early state polls and compare them to national surveys or other states with nearly identical demographics, also I know i'm an Obama supporter but primary's do usually come down to the most motivated to show up, and the donors and crowd numbers can't be ignored here, remember the polls with Tester and Morrison, I think the candidate with the "energy"
behing them has a huge edge in primaries and polling doesn't usually catch it,Howard Dean is the example always givin to discount that theory but his collapse was a very unique set of circumstances, it's hard to see Obama falling apart like that.
by nevadadem 2007-04-14 10:33AM | 0 recs
Re: well
Again, that is just a theory. Clearly, Clinton is not as strong in IA, NH and SC as she is nationally, but the reason for that may not be that those states are more "engaged." In SC, Obama's lead could simply stem from from the very high population of African-Americans in the electorate. In IA, Edwards lead could be seen as an extension from his strong showing in 2004. Or, it could simply be that these statewide polls are actually sampling likely primary voters, rather than all those who self-identify or lean Democratic.

The Iowa poll showing Clinton ahead among "registered Democrats," and showing Edwards ahead among "likely caucus goers," is probably the best evidence yet to support my theory. However, it remains only one data point, and as such is not conclusive.
by Chris Bowers 2007-04-14 10:39AM | 0 recs
Just hear say

I just spoke to a very politically active friend in Iowa and he says that the Clinton campaign does not have the presence that either the Edwards or Obama campaigns have in Iowa.  I found that an interesting observation and a curious fact if true

by Licorice 2007-04-14 10:50AM | 0 recs
it is true

I've gotten personal calls from two different field organizers for Obama. I'm volunteering for the Edwards campaign, so obviously I am getting a lot of communication from them.

I've gotten nothing from the Clinton campaign other than a couple of robocalls advertising upcoming events. Before her big town-hall meeting in Des Moines a couple of months ago, she was running radio ads, but I have not had persona contacts.

Rumor has it (at least on the Iowa Progress blog, www.iowaprogress.com) that the Clinton campaign is having trouble filling her field organizer positions, even though they are offering slightly higher salaries for that job than Edwards or Obama.

I am not working my precinct heavily, but I have started calling some of the more active voters who were at our caucus in January 2004. I have not yet identified any Clinton supporters or leaners. I'm not saying she will have no support in my precinct, but I have not found anyone yet. I've found a lot of undecideds between Edwards or Obama or Richardson. I comment I hear a lot is, "I don't know, but not Hillary" or "I don't know, but I can tell you it won't be Clinton."

This is not scientific proof, but I talk to a lot of active Democrats. There is no doubt in my mind that Clinton has the least support among the most politically engaged voters.

by desmoinesdem 2007-04-14 11:40AM | 0 recs
Re: it is true

anecdottaly I know a few Hillary people in New Jersey, they are not political and simply say Bill was a great president but that they haven't thought much about the race, when I pointed out how difficult it will be for her to win a general election they agreed and are now supporting Obama, I wonder if that explains the difference in polls form the early states and nation wide.

by nevadadem 2007-04-14 01:46PM | 0 recs
Re: well

I strongly suspect your theory ise right, though MP is  right to say the evidence is not conclusive, on the other hand I've argued from Edwards perspective that Clinton being inflated in national polls will be a big detriment to her in the expectations game.

by MassEyesandEars 2007-04-15 03:48AM | 0 recs
National Polls

I think you make some very good points, Chris.

I have little confidence in these polls now, and the issue you identified may make them of little value even when we are closer.  Unless we know that the sample is of real likely Dem primary voters, who knows what we are getting?  What we seem to get is an MSM narrative, perhaps driven not be ideology, but by the desire for a good story in the way Brittany Spears makes good copy.  The Hillobama narrative works for them, because it is new: a black man v. a white woman.  These polls play a role in the narratives, but I think it is partly just a story. If the polls are emprically flawed, it won't matter to them, because they want the story, the horse race and the themes they think will sell.

That is why you and MyDD add so much. Intellectual honesty.  Asking questions that go beyond the chosen narrative of the day.

Good post.      

by littafi 2007-04-14 11:44AM | 0 recs
National Polling Does Inflate Clinton's Advantage

I think that most people who are not Clinton supporters wishing that national polling currently reflects anything more than her name ID and the tremendous amount of free media she recieves would concede that the base of the party is a lot less supportive of Senator Clinton than "registered voters" who are identified as Democrats.

A lot of more casual Democrats feel that they owe her their vote because she is the wife of the last Democratic president and she was attacked by the right wing.  As the campaign becomes more about the issues that will change.  But that isn't the main reason why the polls are so deceptive.

This is an example of common sense trumping pollsters who don't even limit their poll to Democrats who at least say they are going to attend either a primary or a caucus.  Even then, people would say "yes" out of intrigue but at least it would filter more non primary voters out.  They aren't going to be part of the decision making process, so including them in polls, especially so many of them, is bound to wildly skew the results.  

For instance, look at the Iowa American Research Group polling.  John Edwards has won every legitimate Iowa poll since the inception of 2008 Iowa polling (he tied once with Obama in the only poll that included Gore and he tied once with Clinton after her announcement).  Every poll, 12 of them, before and after the ARG polls shows Edwards in the lead.  After the first batch of Iowa polls comes out an ARG poll puts Clinton in the lead, boosting her by 20% points compared to the last non-ARG poll.  More polls put Edwards on top and then the second ARG poll is released and Clinton has an even larger lead.  This poll gives her a 21 point boost compared to the last non ARG before it.  By this time, various comments on Pollster criticize the ARG polls for their lack of a tight enough screen.  Of course more polls come and again Edwards leads.  The first post Vilsack drop out poll by ARG shows Clinton up by 1, a 10 point boost compared to all the other polls, but still far less than the last ARG polls.

You can draw two conclusions from this and neither are good for Clinton.  Either ARG did tighten the screen a little and that is why her lead plummeted in the third ARG poll or they did not tighten the screen at all and this poll simply reflected Vilsack supporters moving to Edwards.  I'm not sure which one it is but it's clear that the polls released since show Edwards in the lead.  In fact, a University of Iowa poll showed the difference between the views of all Iowa Democrats and likely caucus goers.  Of course Clinton led all Democrats but Edwards led the caucus goers.  This is not because of Iowa's exposure to the candidates.  The same principle applies here as well.  More casual supporters back Senator Clinton.

Especially considering that the first two states are caucus states and attending a caucus isn't exactly convenient it is reasonable to conclude that Clinton will lose significant support.  

If there was somehow a magical poll that could see in the future who is realy going to caucus or go to a primary I think that the national polling (as it is today), though bound to change as the campaign changes, considering that support is soft right now, would look something like this...

Clinton 26%
Obama 23%
Edwards 20%
Richardson 5%
Biden 4%
Dodd 2%
Kucinich 2%
Gravel -
Unsure 18%

Even if you disagree with this guestimate I think that as time passes and the views of the Democrats whose opinions actually matters become more clear, there will be no denying that national polling does inflate Clinton's advantage.

by Edwards Supporters United 2007-04-14 12:06PM | 0 recs
Only one data point

But I did a poll on Texas' proposed primary move to Feb 5 that gives a reasonable apples to apples comparison.

The poll first asked if the voter was in favor of the move, then asked if they would vote if it was moved. Third, which primary and finally which candidate.

I ran the identical poll three times using the Texas vote file for a source.
First, I ran it on voters with a history of voting in at least one recent general election.
Second, I ran it on voters who had history of voting in two recent Democratic primary elections
Third, I ran it on voters who had history of voting in two recent Republican primary elections.

Among the general population who indicated they would vote in a Feb 5 Dem primary, Clinton pulled 48%. In the population with a history of Dem primary voting, she was only at 40%. That's a significant difference, though less than a lot of the netroots thought. In doing followup polling for the Burnt Orange Report, I analyzed by region and found that she was very strong in South Texas which votes with disproportionate strength in the primary, while she was neck and neck with Obama in the cities.

by IVR Polls 2007-04-14 12:10PM | 0 recs
the 3 early states demographics combined
are not that different than the nation as a whole
Iowa NH and SCar------the fact that Hillary does much worse in those states can only be atributed
to the increased interest in the race, nothing else can stastically make sense.
by nevadadem 2007-04-14 01:40PM | 0 recs
Re: Do National Polls Inflate Clinton's Lead?

I wonder if Chris Bowers would be conducting such a close analysis of the polls if Hillary weren't the candidate who is leading in all of them.

You won't find any poll that is dead-on accurate, but all of them, when compared over time, provide a reliable trend.

The thing is, if you check the polls from the 2004 campaign, Hillary led most of them, until the polling companies finally removed her name from the list.  Do you suppose all the people who supported her back then have died?  With her success in the Senate, that support is even stronger, and growing.

Hillary has impressive institutional support and a huge base of supporters.  I know it's hard for those who hate her to imagine that.  I also know, just from reading dailykos.com and huffingtonpost.com and mydd.com that there really are human beings who claim they have never met one other human being who supports Hillary.  Well, sorry folks, but that doesn't mean s.h.i.t.
It is just as unrealistic a measure as me saying I have yet to meet one person who supports Barack Obama.  If I repeated that too many times, I would expect someone to tell me to get off my ass and travel the country for a few months.  

Hillary's campaign is just beginning to gear up.

 

by marasaud 2007-04-14 05:34PM | 0 recs
Re: Do National Polls Inflate Clinton's Lead?

I've honestly never had someone tell me that they support Hillary Clinton.

by jallen 2007-04-14 05:51PM | 0 recs
Re: Do National Polls Inflate Clinton's Lead?
"I wonder if Chris Bowers would be conducting such a close analysis of the polls if Hillary weren't the candidate who is leading in all of them."

Ummm.... are you even slightly familiar with the history of my writing? Close analysis of polls is one of my big things, and always has been.
by Chris Bowers 2007-04-14 06:59PM | 0 recs
Re: Do National Polls Inflate Clinton's Lead?

why aren't all the people donating like Obama's supporters are?

by nevadadem 2007-04-14 08:08PM | 0 recs
Barack Obama

I don't believe fundraising came to a stop with the end of the first quarter.  Obama just had a $1000 a plate breakfast; and Hillary has numerous fundraisers planned. The WaPo article, linked below, brings the fundraising results into more of a balance.

Right after Obama's total was made public, there were hundreds of headlines screaming the news that Hillary only took "big money", while Obama made all of his dough in small amounts, from ordinary folks.  The editorial cartoons alone were slamming the hell out of Hillary for being an elitist money-grabber, while depicting Obama as the new grassroots candidate growing support from the bottom up.  

Obama plays it both ways and not just when it comes to fundraising.  As a Democrat, I don't like bashing any of our Dem candidates, but how long are Hillary's supporters supposed to watch the inequities being reported about her and not speak up?  On how many different issues, so far, has Obama been painted as Hillary's polar opposite?  This attempt to manipulate perception is what bothers me the most.

________

From today's edition of the Washington Post (with excerpts below the link).

http://www.washingtonpost.com/wp-dyn/con tent/article/2007/04/14/AR2007041401491. html
"From the outset, Obama tried to establish a "Washington outsider" image -- moving his campaign operations to Chicago and making a bold promise to refuse checks written or gathered by registered federal lobbyists.

The campaign received $50,566 from 49 lobbyists, but aides flagged the checks during initial screening and said they will return the money. Still, for hosting events and otherwise raising money, the Obama fundraising team is relying on partners in lobbying firms who are not registered for specific clients, former lobbyists who recently dropped clients and spouses of lobbyists. The strategy allows Obama's team to reach the wealthy clients of lobbying firms while technically complying with his pledge.

Joanne Hannett, whose husband, Fred, is a lobbyist for UnitedHealth Group and other clients, is helping raise money for Obama. Although Fred Hannett attended an Obama event, he said he has not personally donated any money or "solicited any of my clients."

Obama also has no prohibition against using state lobbyists to raise money, even when they represent companies with business before the federal government."

by marasaud 2007-04-14 10:09PM | 0 recs
Chris Bowers asked:

"Ummm.... are you even slightly familiar with the history of my writing? Close analysis of polls is one of my big things, and always has been."
_________

Yes I am familiar with the history of your writing, as well as your analysis of the polls.

You don't operate with the sheer hatred of Arianna Huffington or Markos, both of whom have become so blatantly biased against Hillary that all they ever acknowledge about her are those things that can be manipulated, or exaggerated, to paint the worst picture possible.  When I found out that one of the co-founders of huffingtonpost.com hosted a high-profile fundraiser for Obama in NYC last week, a little light went on.  Obama's "connections" are not discussed very much are they?  I took particular offense that Obama has Robert Gibbs on his campaign staff, playing a prominent role at that.

Penny Pritzger, Obama's Finance Chairwoman donated to Bush's re-election in 2004.  Michelle Obama's new campaign director, Melissa Winter (hired to help Michelle get more involved in her husband's campaign), has been Joe Lieberman's personal assistant for ten years. Even when these details get scrutinized on the progressive blogs, Obama is given a pass.

I would like all of our candidates to be held to the same standard. If Obama or Edwards should ever make it into first place in the polls, I look forward to reading your diary titled "Do National Polls Really, Really Inflate Obama's (or Edwards') Lead?"   If you are that fair, I will apologize for doubting you.  But the way this diary reads now, it is nothing more than an invitation to cast doubt on Hillary's progress and even, possibly, to suggest that someone, somewhere is "fixing" those polling results. In politics, as you know, a strategic assumption can lead to a rumor that can bring a candidate down.

Let Hillary fail, if that is to be her fate, on her own merits, or lack thereof.  You're a blogger whose commentaries are considered gospel by so many, so you should stick with the facts.  IMO.  Besides, aren't there enough Republican candidates for you to go after?  

by marasaud 2007-04-14 10:38PM | 0 recs
Re: Chris Bowers asked:
So then what was this post? Calling Bullshit On The Blogosphere's National Trial Heat Narrative

I actually can't believe I am not banning you for this comment. Your total lack of respect for the facts, and your willingness to impugne motives where you disagree with conclusions, is disgusting.
by Chris Bowers 2007-04-15 12:35AM | 0 recs
Re: Chris Bowers asked:

I find it interesting that a lot of HRC supporters seem to have a major chip on their shoulders.  I have run into this on a number of sites.  They seem to have a "circle the wagons and shoot on sight" mentality.  

Obama and Edwards supporters are dismissed as being simultaneously hateful and naive.  Perhaps this "tude" is born of frustration that their opponent is not having an easy time of it on the net, or perhaps it says something about the personality type of the folks who are drawn to HRC.

I don't hate HRC, but I think the Dems would be rock stupid to nominate a candidate who is  vulnerable on so many fronts when we have two excellent alternatives in Obama and Edwards.

by upper left 2007-04-15 07:42AM | 0 recs
If hillary wasn't leading

none of this would be necessary.

but she is.  so the polls must be wrong.

by Stewieeeee 2007-04-14 11:30PM | 0 recs
If Hillary wasn't leading

none of this would be necessary.

but she is.  so the polls must be wrong.

by Stewieeeee 2007-04-14 11:30PM | 0 recs
To those whom I gave one's
You might want to read this post. It is back when I argued polls were inaccurately showing Lamont too far ahead of Lieberman pre-primary. Again, I was right. Of course, I took endless flack for arguing it at the time.

Seriously, impugne my integrity on this, and you will shortly find yourselves banned. You are the ones who can't get past your own biases. My record on this front is unimpeachable.
by Chris Bowers 2007-04-15 12:23AM | 0 recs
I'll respond to that

cause i just acknowledged the warning.

Maybe if the invective against hillary wasn't pitched at such an angle across the entire blogosphere.  if the movement as a whole wasn't so defiantly against one particular candidate.. if the mission weren't, so to speak, to defeat the clintons, if there was never any tendency to pile on, ... then such an analysis from a blogosphere icon such as yourself would not be viewed through that lens.

having read the lamont piece, i have a better perspective, but i still regard the notion of impartiality -- most especially during this primary -- very skeptically.  cause if one wants clinton to lose, one is simply more inclined to ask themselves, when seeing a poll showing clinton with a big lead:  "well.   is that right?"  "is there something wrong with the poll?"

and if edwards or richardson showed up with a huge lead -- for whatever reason -- hey.  why look a gift horse in the mouth?  that's good momentum.  

that's a less snarky way to put my comment above.  if clinton was suddenly trailing by 10 in the polls there is less tendency to question the polls.

let me take this one step further.  james carville was attacked for not being labelled a "Clinton supporter" by cnn.  it became necessary for CNN to acknowledge the fact that Carville wants hillary to win and force all of his analysis to be viewed through that lens.

i think you see where i'm going with this.  and while it might be different because carville is on a national broadcast and you're just on a blog, the end result is the same, rendering carville's analysis (impugning his integrity?) to nil except viewed through the lens of his partiality.

what's funny though, carville was criticizing obama's performance at the health care forum in Nevada.  which was actually a valid criticism.  

turns out, maybe your analysis above is equally valid.  

by Stewieeeee 2007-04-15 08:44AM | 0 recs
Re: Do National Polls

"Seriously, impugne my integrity on this, and you will shortly find yourselves banned" (Chris Bowers)

"I actually can't believe I am not banning you for this comment. Your total lack of respect for the facts" (Chris Bowers)
__________

Challenge Chris Bowers and you get banned.  I see how this works.

How powerful this must make Chris Bowers feel; and how very Republican of him.

by samueldem 2007-04-15 05:37PM | 0 recs

Diaries

Advertise Blogads