« Fox Poll Shows Kerry Lead in Battleground States | Main | Kerry Takes Lead in New National Poll »

Why the Race Is Closer Than People Think

Is Bush ahead by a little or a lot? Is it close to a tie ball game or has Bush surged to a commanding lead?

The conventional wisdom inclines to the latter not the former. The reason has a great deal to do with two persistent problems with contemporary polls that--at least at this point in time--tend to considerably inflate Bush's apparent lead. But once you dissect the available data with these problems in mind, a truer picture of the race comes into focus which suggests that the race continues to be very close.

The two problems are: (1) samples that have an unrealistic number of Republican identifiers and hence tend to favor Bush; and (2) the widespread and highly questionable practice of using likely voters (LVs) instead of registered voters (RVs) to measure voter sentiment this far before the election.

First, the issue of partisan distribution in samples. Lately, and very suddenly, many polls have been turning up more Republican identifiers than Democratic identifiers in their samples--in some cases, many more (as high as a 9-10 point Republican advantage).

How realistic is it to be suddenly turning up a Republican lead on party ID, much less a large one? Not very. The weight of the academic evidence is that, while the distribution of party ID among voters can and does change over time, it changes slowly, not in big lurches from week to week.

And the weight of the empirical evidence is that the distribution of party ID among voters has favored and continues to favor the Democrats. In 2000, the exit polls showed Democrats with a 4 point advantage over Republicans. In 1996, it was also 5 points; in 1996, it was 3 points and in 1988 it was also 3 points.

The data also indicate that there were two shifts in party ID over the 2001-04 period which largely cancelled each other out. The first shift, in the period after 9/11, shaved several points off the Democrats' lead and brought the Republicans close to even (but never ahead) in party ID. The second shift tooks place in late 2003 and 2004 and reconstituted the Democrats' lead on party ID to about 4 points, exactly where it was in the 2000 election according to the exit polls (see this useful study "Democrats Gain Edge in Party Identification" by the Pew Research Center for more details).

So, if polls are suddenly turning up too many Republican identifiers and that is unrealistic and skews reported horse race results toward Bush, what, if anything, should be done?

One possible solution is to weight poll results by a more reasonable distribution of party ID. The issue of whether to use this approach to the problem is well-summarized by Alan Reifman in his invaluable essay "Weighting Pre-Election Polls for Party Composition: Should Pollsters Do It or Not?" on his website.

As Reifman puts it:

One factor (among many) that may contribute to discrepancies between different outfits' polls in their Bush-Kerry margins....is polling firms' different philosophies as to whether it's advisable to mathematically adjust their samples -- after all the interviews have been completed -- to make the percentages of D's and R's in their survey sample match the partisan composition that is likely to be evident at the polls on Election Day. The latter can be estimated from exit polls from previous elections, party registration figures (in states where citizens declare a party ID when registering to vote), and surveys.

(Another issue that often comes up in evaluating pre-election surveys, with which many of you may be familiar, is whether results are reported for "registered" or "likely" voters. That is a different issue from what is being discussed [in this essay]. Whether a pollster reports results for registered voters, likely voters, or both, weighting by party ID is a separate, independent decision.)

Note well Reifman's point that the issue of whether and how to use LVs, not RVs, to report results is separate from the issue of whether and how to do party-weighting. I discuss the LV issue below after the party-weighting discussion.

Given that party ID does shift some over time, my instinct has generally been to avoid party-weighting if possible and promote a full-disclosure approach. This is how I put it in a recent post:

[B]ecause the distribution of party ID does shift some over time....polls should be able to capture this. What I do favor is release and prominent display of sample compositions by party ID, as well as basic demographics, whenever a poll comes out. Consumers of poll data should not have to ferret out this information from obscure places--it should be given out-front by the polling organizations or sponsors themselves. Then people can use this information to make judgements about whether and to what extent they find the results of the poll plausible.

But this approach increasingly seems unrealistic to me. The polling organizations and sponsors do not routinely release the data I call for and certainly do not prominently display them. And even if they did, the typical consumer of polling data lacks the time and skills to use these data to re-weight or adjust reported results. The fact of the matter is that people pay attention to reported results period; therefore they are at the mercy of whichever results are reported and emphasized (an issue that also looms large in the LVs vs. RVs issue, discussed below).

This suggests that weighting poll results by a reasonable distribution of party ID may be necessary to avoid giving the public distorted impressions of the state of the race.

What is a reasonable distribution of party ID to use in such weighting? One obvious candidate is the exit poll distribution from 2000: 39D/35R/26I. Moreover, the Democratic advantage in this distribution--4 points--closely matches the average Democratic advantage in 2004, as measured by the Pew Research Center (see above) and other polling organizations, making it an even more attractive option.

But political analyst Charlie Cook probably has the best idea, even though it can really only be implemented by the polling organizations themselves: "dynamic party identification weighting". Cook's idea is that polls should weight their samples by a rolling average of their unweighted party ID numbers taken over the previous several months. This would allow the distribution of party ID to change some over time, but eliminate the effects of sudden spikes in partisan identifiers in samples (such as we are experiencing now).

Lacking such a dynamic weighting, however, the best we can probably do at this point is to use the exit poll distribution mentioned above. How much difference would this make if we applied it to recent polls?

Quite a bit. Here are Bush's leads in a number of recent polls, ordered by size of his lead, once the horse race question is weighted by the 2000 exit poll distribution (note: not all recent polls can be included because you need the horse race figures among Democrats, Republicans and independents separately to do this procedure and not all polls release these figures; in addition Zogby and Rasmussen results are party-weighted to begin with and therefore do not have to be re-weighted; RV results used unless only LV results available):

CBS News, September 6-8 RVs: +5
Zogby, September 8-9 LVs: +2
Rasmussen: September 10-12 LVs: +1
Fox News: September 7-8 LVs: +1
Washington Post, September 6-8 RVs: +1
Newsweek, September 9-10 RVs, -2
Gallup, September 3-5 RVs: -4

These data present a clear picture of a tight race, with Bush likely running a small lead, but not the solid--and even large--advantage that has been conveyed to the public.

The other problem that is afflicting the polls and considerably inflating perceptions of Bush's lead is the widespread, and highly questionable, use of LVs, instead of RVs, to report horse race results far in advance of the actual election. The reason why using LVs instead of RVs is a bad idea is simple: the LV approach is being asked to do a job--gauge voter sentiment and how it changes from week-to-week (and even day-to-day)--that it was never designed to do. What the LV approach was designed to do was measure voter sentiment on the eve of an election and predict the outcome. That was, and remains, an appropriate application of the LV approach.

But applied as many polling organizations currently do, it is highly inappropriate and frequently very misleading. As political scientists Robert Erikson, Costas Panagopoulos and Christopher Wlezien put in in their important forthcoming paper, "Likely (and Unlikely) Voters and the Assessment of Campaign Dynamics" in Public Opinion Quarterly:

[E]stimates of who may be likely voters in the weeks and months prior to Election Day in large part reflect transient political interest on the day of the poll, which might have little bearing on voter interests on the day of the election. Likely voters early in the campaign do not necessarily represent likely voters on Election Day. Early likely voter samples might well represent the pool of potential voters sufficiently excited to vote if a snap election were to be called on the day of the poll. But these are not necessarily the same people motivated to vote on Election Day.

And of course, since the group of people "sufficiently excited to vote if a snap election were to be called on the day of the poll" changes from poll to poll, it raises the uncomfortable possibility that observed changes in the sentiments of "likely voters" represent not actual changes in voter sentiment, but rather changes in the composition of likely voter samples as political enthusiasm waxes and wanes among the different parties' supporters. Or, as Erikson et. al. put it:

At one time, Democratic voters may be excited and therefore appear more likely to vote than usual. The next period the Republicans may appear more excited and eager to vote. As Gallupís likely voter screen absorbs these signals of partisan energy, the party with the surging interest gains in the likely-voter vote. As compensation, the party with sagging interest must decline in the likely-voter totals.

And this is exactly what their analysis of Gallup data from the 2000 election finds--"shifts in voter classification as likely or unlikely account for more observed change in the preferences of likely voters than do actual changes in votersí candidate preferences".

This is an important result and helps nail down what has always been disturbing about the use of likely voter methods far in advance of the actual election. Instead of giving you a better picture of voter sentiment and how it is changing than conventional RV data, it gives you a worse one since true changes in voter sentiment are swamped by changes in who is classified as a likely voter.

Does this matter? You bet it does. When Gallup told the world on September 6 that Bush was leading Kerry by 7 points among LVs, the world listened and absorbed that figure as a trustworthy indicator of where the race was. Completely lost, except to those who bother to look at such things, was the Gallup finding that Bush only led by single point among RVs--in other words, that the race was about tied. Gallup and its sponsoring organizations implicitly and explicitly encouraged people to treat the LV finding as the real story and the RV finding as an unreliable afterthought (after all, those voters aren't "likely"!). The incredible irony, of course, is that the real situation was exactly the reverse: as the Erikson et. al. findings suggest, it was the RV data that provided the best gauge of voter sentiment and the LV data that should have been an unreliable afterthought.

Or take the Gallup data gathered in Ohio in the last two months, perhaps the key state in this election and the subject of endless media stories about "the battle for Ohio". On September 8, Gallup released data showing Bush ahead of Kerry by 8 points among LVs in Ohio, a 14 swing from late July when Kerry led by 6. Again, completely lost in the Gallup, newspaper and television reports on the poll was the poll's finding that Bush had just a 1 point lead among RVs in the state, representing a much more modest swing of 6 points since late July.

Guess which figures are still with us as coverage of the battle for Ohio continues? That's right: Bush's 8 point lead among LVs and 14 point swing. In fact, just this Sunday, The New York Times practically built their Ohio campaign story around these figures which showed just how well Bush is doing! and just how much the situation has changed!.

In short, these LV figures, especially from Gallup, are contributing mightily to the impression that Bush has built a substantial lead and is even surging ahead in some of the key swing states. But, as we have seen, these LV data are fundamentally inappropriate for measuring the state of the race, and how it is changing, this far ahead of election day. For that, you need the RV data and they suggest something far different: the race is damn close and Bush's substantial lead is a myth.



Bravo! I am an engineer who knows about polls and statistics, and have been worried that the Gallup (and Time) polls have the potential of becoming self-fulfilling prophecies. Why do they do it? I presume that they are not hacks who must read the same literature you do, and thus know what is right and what isn't. Yet they persist in using the suspect methodology. Again, why? Can you shed some light on this? Venture a guess or hypothesis?

Even worse are the innumerate political reporters such as R. W. Apple of the NYT who reported the egregious Ohio story in Sunday's Times. CNN is one of the worst offenders, because they tend to parrot only the numbers with the biggest margin (e.g., Time's 11-point poll). What can be done about this? Require training in polls for all reporters? (Excluding Fox and the like for obvious reasons.) It would seem that this is the only way to remedy this, only it won't happen this election.

Keep up the good work. Yours is a voice of reason in a cacophony of cant.

Raul Martinez

Frequent reader. First time poster. I really love this site. And I found this post very informative. One question, though. Much of your analysis is predicated on the notion that for some reason a number of recent polls have turned up an inordinate number of Republican identifiers. Any thoughts on why that might be? It seems odd that a variety of polls over the course of a couple of weeks would err in this same way.



Thanks Ruy for the info, my question is simular to Raul's, can't we insist or protest or teach these reporters to properly report these polls? Especially if the pollsters themselves do not?? It seems like a collective effort is needed to forse this issue. Can you start some sort of campaign to report the data properly??

Telephone polls get a disproportionate response because old fart Republicans will answer the phone and talk to anyone, while cooler Dems don't answer any fool who happens to hit their digits.

Gabby is right, in a flippant sort of way. Telephone polling could be increasingly skewed as telecommunications technology and telemarketing have changed the way people react to unsolicited calling. A lot of people in metropolitan areas for example are home infrequently or just don't answer a call they can't ID. On the other hand internet polling has become more and more interesting. The recent globe article suggests this could be far more accurate that telephone polling.

Great analysis. Its seemed like, to me, that this amazing lead Bush has enjoyed overthe past few weeks is largely fictional. This goes a long ways in explaining it.

My question is: will this effect voter turnout or possibly change voter's opinions on Election Day?

I'm not right in a flippant sort of way. I'm right in a blunt sort of way.

Can you imagine wanting an accurate marketing poll and getting one of these surveys?

You'd ask "where the hell are all the young people who don't vote Republican?"

The answer would be "not answering their phones to the pollster."

Polling people by phone or online is inherently flawed.

this is getting ridiculous. i want to believe kerry is not losing too, and perhaps he isn't. but with every poll that comes out, ruy comes up with some explanation of why it doesn't mean what it apparently means. it seems to me this is not the first time in a recent election that democrats have looked at polls and said "everything is fine!" and found that election day shows those polls to have OVERestimated democratic turnout, not underestimated it. why is this year different?

the question is: what evidence would convince ruy that kerry is actually losing? if no actual evidence can change your theory, that's not a theory, that's blind faith. and while some faith in hard times could help us win this election, at some point we might need to realize we need to change direction before it's too late.

Hold on there.

You are presuming attacking the accuracy of polls is the same as giving a pass to Kerry on poorly advancing his cause.

We have been all over his ass about not doing a better job. Here, I'll list some quick do's and don't for Kerry:


1. Have Edwards open for you, deliver the message
2. Have surrogates savage Bush et al.
3. Hammer sound bite themes such as
**Four years ago ....
**George Bush has FAILED ....
**Take America BACK from bad leadership ....
4. Roll your sleeves UP
5. Talk less, say more
7. Up and Down:
JOBS down. INCOME down. OIL profits UP.


1. Windsurf or cycle again
2. Let THK speak
3. Wear a friggin tie so much
4. Use such long sentences
5. Talk AT people. Life is not a speech. Converse.

One other thing is that the pollsters really don't care about a poll 5-8 weeks out from election day. Their is a "polling sweepstakes", in the final poll before election day, every polling org wants to be the most accurate for that one. (BTW-in the last two elections Zogby has won that mythical award).

I see the pollsters as currently "tinkering" with their panels, and models. But in the meantime damage is being done. All this "bad news" for Kerry cannot help.

I wish something could be done.

I'm afraid that I'm still not convinced. I can see why Republicans would be overrepresented in telephone samples, but not why that effect would only start showing up recently. Ruy, am I right that that's what's happened and, if so, why do you think it might be?

I'm tired of people saying Theresa should be hid in the basement like some kind of embarassment, just because she has strong opinions and the guts to speak out.

I'm sick of how women in general are marginalized in this country, even a billionaire woman like Theresa. What does that say for this country, even a woman who has a billion dollars is gagged and hidden like an embarassment.

I think Theresa is onto something when she said that anyone who disagreed with Kerry's health plan is an IDIOT. Because that's how I feel too. Let's run campaign commercials that say exactly that, here is our health plan, if you disagree, you are an IDIOT. And NO SEX FOR REPUBLICANS. We will win in a landslide. Women united will never be defeated!

Good advice.

1. Windsurf or cycle again
2. Let THK speak
3. Wear a friggin tie so much
4. Use such long sentences
5. Talk AT people. Life is not a speech. Converse.

I am still amazed that he doesn't yet have a concise, engaging speech or well-rehearsed talking points.

If you are serious about running for the Presidency put some effort into learning how to give a good speech. I don't want to conclude that it is just laziness, but I am running out of reasons to explain it. Is it too late to flip the ticket? Just kiddin' - Kerry wins.

Women united will never be defeated!

Posted by susan at September 14, 2004 02:45 AM

The point is to win the election, not satisfy your need for gender validation.

More loser thinking from the left.

If THK had great positives and low negatives, I'd say use her. But she doesn't. Use Edwards (wife not Senator), or Boxer, or Feinstein, or any number of legitimate political commentators who are female but not loose cannons with high negatives.

We are electing a candidate here, not filling out your personal dance card. It's not sexist to suppress negatives. It's sexist that she got the negatives through a concerted effort, but we can't fight that battle. Keep her out of sight.

I am new to this list and like most people here find the selective use and over emphasis of certain polls to be highly damaging to the Kerry campaign. This is done with disturbing regularity on CNN and MSNBC.

However, I do have a question. The latest poll shown on pollingreport.com shows Kerry-Bush tied among likely voters at 47/47 and also has Kerry ahead by 2 among registered voters. This is a IBD/CSM/TIPP poll that ended on 9/12. Does anyone on the list know the historical reliability of this poll? I am wondering if it is showing any new trend.


((I can see why Republicans would be overrepresented in telephone samples, but not why that effect would only start showing up recently.))

Reread his article. Ruy doesn't claim that the effects of bad weighting and LV screens started "showing up recently."

((It seems to me this is not the first time in a recent election that democrats have looked at polls and said "everything is fine!" and found that election day shows those polls to have OVERestimated democratic turnout, not underestimated it.))

As has been stated here before, in 2000 many major polls, most prominently Gallup, reported Bush up by 3-4% while he actually received -0.5%. It's not as if little precedent exists for systematic errors like those Ruy cites. As for WHY organizations like Gallup who have no shortage of PhDs on staff allow such biases to go unaddressed... now that's a good question..

It seems to me as if JFK's big problem has been getting a consistent, sincere message across to voters. I think his obsession with seeming "tough" occasionally fools him into supporting this Administration's bad policies. For example, "Shrub" recently made a blunder by actually telling the truth about lacklustre chances of the U.S. winning the War on Terror using military means. So JFK immediately attacks his "lack of resolve" or whatever. Now some bloggers are urging JFK to accuse the President of cowardice because the U.S. forces have effectively ceded Fallujah and several other cities to the insurgents! Don't they understand how pretentious this sounds since JFK always has been very cautious and sceptical about this sort of thing? And for a very good reason too because *he* was right about this whereas "Shrub" & co. were dead wrong all along!!! Yet much of the time he seems to be trying to attack the President from the _right_ on Iraq! What's the point? The Republicans will *still* smear him as a weak liberal flip-flopper!

In Iraq, there are basically two tough options. If the U.S. wants to follow the course set by this Administration, it needs to deploy some 400,000-500,000 troops and annihilate the insurgents in every city while finally providing adequate security for ordinary Iraqis as well as foreign aid workers and contractors trying to rebuild the country. Only then will things get better, but it will be costly in terms of casualties and the financial expenses will be staggering too. The alternative strategy is to recognize the current (relatively limited-) troop presence is basically useless since it cannot maintain law and order anyway, and because it basically encourages anti-American elements to keep killing people. Kerry should show some backbone by pointing out this inconvenient fact to voters. And it's usually the same story no matter what "Shrub" policy you look at: he simply has never been willing to pay for anything or make hard choices. He still thinks we can afford record tax cuts at the same time as the scope of the federal government is greatly expanded. The thought the war in Iraq would cost nothing, so he refused to pay for that too, or to deploy enough troops.

The truth does hurt, but at least Kerry would lose respectfully if he presented voters with the tough, hard choices (there *is* a war going on, after all). He could say he honestly thinks all U.S. troops should be brought home within the next 1-2 years no matter what happens, because they will do little good unless the U.S. is willing to devote much more money and manpower to the war there. He should also urge the current President to support *his* own policy by boosting the Iraqi military presence by hundreds of thousands of soldiers, if he really thinks the current strategy needs to be pursued after the elections as well. _That_ is the real lesson of Vietnam: if you fight a war, you set aside adequate resources and make sure the "home front" stands united behind the soldiers.

If Kerry had been consistently pushing this foreign policy theme during the last year or so, the worst thing that could happen is he would lose after being right on the issues. He would also receive a record number of votes from disgruntled fiscal conservatives.


I think that no matter the real truth about the polls and their weighting, the results can only do two things to Kerry supporters. They will either get depressed and sulk and whine and complain, or they will simply be motivated to ensure that the polls were wrong on Nov 3.

Its a decision that both the Kerry camp and the Kerry supporters have to make. In the final analysis, its not about the inner workings of the polls but who is motivated enough to trot pass the tape and win.

Personally, I am not disappointed that the polls try really hard to give bush a commanding lead because I am hoping that this motivates people enough to ensure that they crawl or creep to the booths and vote for Kerry. I hope these polls and the bush media gets people mad as hell and hence they re-double their efforts to get people registered and then drag them to the booths to vote for Kerry.

So in effect, its really a matter of what effect the kerry supporters allow the polls to have on their psychy. I hope its the "let us send bush packing" effect.

The analyses of the inner workings of the polls are quite interesting however. I enjoy reading them too. In any event, I am still very convinced that Kerry will defy the polls and those who think his message is weak, waffling and in some instances non exisitant, and win on Nov 2.

I am pretty sure that I am not alone in my convictions. Raise a hand if you agree with me.


This is article is worth highlighting and bringing to national attention....


Once again Ruy. WHEN do we start looking at LVs? Isn't it time YET?

CNN poll the day before the 200 election was Bush 48, Gore 43, Nader 4.

The last posting by Bel brings up a very important point. Do we have any empirical evidence that the stated results of a poll have any effect on the results of future polls? If not, then all this discussion of unrealistic party affiliation and LVs vs. RVs is, while interesting, is only of academic interest. As Bel indicates, disappointing poll results could depress the followers of a politician OR motivate them even more. It is not clear which result will follow. It would be interested to learn whether there have been any studies done that address this issue.

Here's a question about polling samples that I have not heard addressed. Given the proliferation of cellular phone use, particularly since the 2000 election, aren't these telephone surveys even more unreliable? There is no cellphone directory, and even if pollsters had cellphone #s I believe it's still illegal to call them for solicitation, including polling. So it would seem to me the "wireless" generation is vastly under-represented in polls that simply rely on someone with a traditional wall phone to answer it. To me, this would seem to skew the results dramatically. Any thoughts???

I agree. People giving up their landlines for cell phones is a huge story. It's the elephant in the room that pollsters seem to want to ignore. In my view there are strong reasons to believe that this group skews Democratic:

1. Urban voters are the most likely to give up their landlines and rural voters are the least likely, for obvious reasons: lack of cell phone coverage in many rural areas.

2. Young people are more likely than older voters to embrace new technology.

3. Blacks, according to one Michigan survey, are three times more likely than whites to give up their landlines for cells.

4. Low income people are likely to favor "pay-as-you-go" cell phone plans over fixed monthly charges.

I'm phone banking for Kerry in some low-income wards, and the number of disconnected numbers is incredible. My hunch is that a large chunk of these are people who have given up their landlines for cells.

The pollsters themselves know the problem exists, but don't know what to do about it. "We're in a muddle trying to understand what to do," said one.

What do you think Ruy?

My theory on why the numbers of self-identified Rethuglicans have increased in recent polls: the Swift Boat Veterans smear campaign. This red-meat energized the GOP base, and probably had the opposite effect on many Democratic voters.

I'm no expert, but in the past few months I looked at the websites of a couple of polls and found a question way down at the end: "How did you vote in 2000?" and both of them had Bush about 7% ahead of Gore! If this isn't proof that they oversampled Republicans I don't know what is. (Unless Bush really won after all, and it was the Democrats who were guilty of massive vote fraud; which I strongly doubt...)

good analysis. But to be honest, if polling organizations dont change their methodology, then these polls are very useful for trending. Its way to early to be using them as snap shots effectively. And CLEARLY the trend isnt positive for kerry.

Kerry lacks a focus. HE spent a good week bashing on about "Wrong" but the last few days he is all over the map again, just reacting to bad news about every issue. Assualt weapons, NK, medicare etc.

And to be honest, he doesnt have a coherent message for Iraq yet. I am hoping he is going to time all this for the debates, but if they come and go, then the field is set and its the hand that has to be played

Am I the only one who's seen the WSJ poll (online edition of WSJ) of the battleground states? It's sourced from an interactive Zogby poll, but it shows Kerry leading in almost all battleground states. It can be found under the WSJ's Politics and Policy section, Campaign 2004.

This is excellent and the kind of explication I've needed badly. I haven't had a course in college level stat since 1958, but recall enough to be dangerous. I've been following politics since Eisenhower, years before I was old enough to vote. Gut feel is no substitute for hard data, but my gut tells me that this election will be a mirror of 2000. If that's correct, then these numbers showing Bush way ahead are likely artifactual and not "real". Bush's net negatives simply contradict the horserace numbers.
This could turn out to be another 1936 Literary Digest fiasco. It might be therapeutic if it did.

I think the basic question raised by the first
part of this post (and not addressed in this interesting discussion) is *why* there has been
an apparent surge in Republican leaners
in samples taken by various leading polling

I'm worried about something besides the polls. It's what Raul at the top mentioned.

The media have latched strongly onto the narrative that Kerry is losing the election. Discussions about the election morph into post mortems (really, pre-mortems) about how Kerry lost, about what he's doing badly now, always with the clear assumption that he's already LOST.

This *can and will* have a self-fulfilling prophecy effect.

I'm emailing and complaining to cnn and msnbc more than once a day about this. Maybe if others join in, we can get some fairness here.

It amazes me how the Bush machine is able to direct and control the media so effectively. It's like we're living in a totalitarian state. You add this on top of the way they've kept the news of new casualties off the air during an election season, and the way they got the SBVT attacks to saturate the TV, and the way they have now -- the latest thing -- got Kitty Kelly's book frozen out of all the major talk shows except for the Hardball. Where she can expect a cold reception -- my prediction. Also, remember how they almost succeeded at killing Fahrenheit 911, to the point of intimidating Disney into not distributing it, and then intimidating theaters that showed it (many towns never got to see it because theaters refused).

We need to do something about this. Somebody brighter and more effective than me has to. Where is the Democratic party operation when we need it? Terry Macauliffe may be a great money raiser, but if he can't organize the party to fight back more effectively in the media, then I wonder what good he is. All the TV ads in the world won't help us if Bush can effectively control the news.

pdb: "How did you vote in 2000?" and both of them had Bush about 7% ahead of Gore! If this isn't proof that they oversampled Republicans I don't know what is"

Actually, I would attribute some of that to as re-identifying with the winner...

I'm obviously not an expert but I've heard counter-claims to the importance of party identification. One writer cited a spike of Republican-identifiers just before the the 2002 election. Is it possible that party identification is an indicator (perhaps a lagging indicator) of candidate preference. In other words, if I have decided to vote for Bush I may re-evaluate my status as a Democrat or Independent.

Great analysis. I think we all need to breath and understand that the next week will see a tightening of the race, a switch in the talking head to a "tighter race" and the question of "Is Kerry turning things around"...Plus, I have faith that the debates will be a plus for Kerry.

I do wonder about the surveys that claim that significantly more people voted for Bush than Gore. Either the samples are unrepresentative or people are lying (perhaps Bush converted them a while ago and they forgot that they voted for Gore.) I think it's a bit of both.

I'm not going to go into a detailed analysis of each polling result, but I think the media should provide an average of what all reputable polls say, not just pick one result (often an outlier) and treat it as gospel. And don't forget about margins of error; a three or four point gain is not a "surge".

Also, it's become a cliche to say this, but polls represent a "snapshot in time", not a reliable predictor of what will happen weeks from now. Trying to identify likely voters is guesswork, and is subject to errors and biases (and is often proven wrong on election day.) For Democrats, the trick is not to let bad poll results seven weeks out discourage them from voting on election day. If we are discouraged and don't vote, then we get what we deserve.

And I was wondering, is USA Today/Gallup restarting its daily tracking poll? In 2000 they gave us daily updates and they ranged from double-digit Bush leads to double-digit Gore leads, sometimes within the same week (all of which were hyped as surges for one candidate and big trouble for the other.) I was thoroughly confused (kind of like now.) The only daily one I'm really aware of is Rasmussen, which has so far been more stable.

I am extremely nervous about the media's misrepresentation of the polling data. They are predicting the outcome they want, not the outcome I still think we are going to get.

Some democracy: first the voters in two of the most conservative, hardbed Republican states pick our Democratic candidate, and then the media counts him out 50 days before the election. It's all over? Hey, I'm still sitting here at the Jersey shore waiting for election day. Nobody's asked me my opinion yet, and I guarantee there are hundreds of thousands of people like me who are paying attention to what matters in this election. And come November, Gallup, Newsweek, Zogby and the rest will find out what I think.

Dumbo, I also e-mail CNN,MSNBC, NBC, ABC, nothing does any good because look to where these news casters and there copy writers are getting there pay checks. General Electric owns NBC and GE makes most of the ammunition for the war. Time Warner owns CNN Time Magazine. I think CBS,Viacom, has more freedom to speek the truth now days..............

I forgot Disney owns ABC. The press has freedom as long as they are shills for the right wing. Ted Turner sold CNN to Time Warner and who came out with the first poll after the convention?

jonathan wrote:

"As has been stated here before, in 2000 many major polls, most prominently Gallup, reported Bush up by 3-4% while he actually received -0.5%. It's not as if little precedent exists for systematic errors like those Ruy cites. As for WHY organizations like Gallup who have no shortage of PhDs on staff allow such biases to go unaddressed... now that's a good question.."


Gallup picked Bush by 2.5 million voters and he lost by .5 million. That is exhibit number one in the trial of Gallup's presumed bias in its LIKELY VOTERS.

No one ever starts a story with "for what it is worth, a poll that missed the election badly last time is once again saying that Bush is several million voters ahead."


Dems voting for Kerry 41%, against Bush 51%
Repubs voting for Bush 82%, against Kerry 11%

Did you see the tape of Kerry speaking before the Congressional Black Caucus. Other than Al Sharpton standing and applauding in the front, most attendees looked very apathetic. I don't think you guys are going to turn out the Black vote this time. They ain't feeling Kerry.

I'm pretty sure the Jewish vote is looking for the exits as well.

Not good.

4 More Years!

Hey All,

In a effort to shift the discussion -

I attended a MoveOn PAC "Leave No Voter Behind" organizational meeting. It was inspirational. The room was full to overflow. There is need for a second meeting on Thursday. I am a life-long Democrat and I have not seen this level of commitment/angst ever!

Also, as I have mentioned before I believe and many conversations support my belief that there is a significant portion of the electorate who plan to vote for Kerry. Perhaps they are not in love with Kerry, however, they will, over time, come to appreciate his ability to make nuanced responses and balanced CONSTITUTIOAL decisions.

I hope you are all as actively involved in voter registration and get out the vote efforts. The larger the turn out the greater our margin of victory.

Oh yeah - Rasmussen has Bush ahead less than one percentage point and has Kerry ahead by nearly five percentage points in the 16 battleground states.


BJ, You can count me as one of the people that's voting against Bush more than voting for Kerry. I don't see the point, honestly, in your making such a big deal about it. I want to get rid of Bush so badly, I'd vote for just about anybody to replace him.

Hell, I'd vote for any Republican to replace him, too, if that was my only choice. I was a hard core hawk Republican for many years (80's, early 90's). I think Bush is not just a crappy president; he's doing enormous destruction to this country and has to be removed by any means necessary. So, yeah, that weighs more heavily in my choice making than the fact that John Kerry is one cool guy, and a hero, too.

"Dumbo, I also e-mail CNN,MSNBC, NBC, ABC, nothing does any good because look to where these news casters and there copy writers are getting there pay checks. General Electric owns NBC and GE makes most of the ammunition for the war. Time Warner owns CNN Time Magazine. I think CBS,Viacom, has more freedom to speek the truth now days.............."

ABC News Now is looking very good.

Keep emailing those guys. I know the email gets trashed, but I keep doing it for a reason. I visualize a situation where two producers are arguing about whether or not to go with an anti-bush or pro-kerry story, and the one says to the other, "What about all the email accusing us of being pro-Bush?"

By emailing them, even if they don't correct a mistake they made, it gives some ammo to the more fair reporters/producers in the news room. That's my little mental justification, anyway. So please, keep it up. You and me. :)

And in an effort to shift the discussion back onto the topic...

It actually, doesn't matter all that much which data you look at or who's data it is. If you look at enough data from different sources, you get pretty much the same answer any way you look at it.

If you take the average of all the published national head-to-head polls done August 1-15, giving preference to LV sub-samples and 3-way race numbers wherever available, you can see Kerry coming out of the Democratic convention with a lead of about 3-4%. If you give preference instead to RV, 2-way data you get a Kerry lead of more like 4-5%.

If you buy my methodology, then the presumed MoE's on the aggregate sample are going to be in the neignborhood of +/- 1%, so while the Kerry leads are probably significant, the differences between them in these two scenarios are not.

Now if you look at the numbers so far for the first half of September, coming out of the RNC and perform the same exercise, you see Bush piling on a 5-6% lead if you look at his best case average. And again, if If you give preference instead to RV, 2-way data where available instead, that's more like 4-5%. And if you take the latter and throw out the three biggest outlyers (because Damn!) it's Bush by 3% (48/45).

So my contention is that both candidates got their post-convention bounces but both were difficult to make any sense of because both were right at the edge of the MoE -- just as I was preaching to anyone who would listen back in June. It also doesn't help that with ~90% of the electorate having already made up their minds about whom they intend to vote for, the only place left to get any bounce is among the most volatile segment of the electorate.

So there's an awful lot of jitter in all the polls right now, but averaging does serve to reduce the effective noise level somewhat. And right now, it looks like the number 46 is beginning to re-exert it's magical attraction on both candidates, as we have been observing all year. So I'm really expecting the slug-fest to continue right through election day.

For a completely different prediction method, check out the Iowa Political Markets site (http://www.biz.uiowa.edu/iem/markets/). The idea is for traders to buy futures on election results based upon what they think will happen.

Right now, people are buying George Bush. Time to buy Kerry!

Thank you for this analysis. My dad keeps on calling me telling me that pollsters ought to be dragged out and arrested. I just sent him this article to calm him down!

After reading these posts, it seems that this is really an issue that calls out for a methodical study to answer a very basic question: will a thousand numbers pulled at random from telephone books nationwide yield an accurate sample of the American voting population? Someone could call cellphone companies and ask them about the demographics of their customers--are they disproportionately urban and African-American? In 2000, it seems that the polls overestimated the extent Republican support--but the failure of polls to accurately predict the outcome of the election never materialized as an issue because it was swamped by the imbroglio in Florida.

Is anyone aware of any media watchdog organizations that have taken on the issue of standards for which polls to report, and guidelines for how to report them? I know of individuals--several here, for example--who are making noise on this.

Posts like this by Ruy help make it more possible for individuals and organizations to know enough to be able to raise the sorts of questions any newspaper editor with even a shred of integrity would take seriously.

I share the interest of the poster who inquired what reputable studies have shown on the effects, if any, of reports of polls on subsequent citizen campaign and/or voting-related behavior. If there is compelling evidence that it has *no* effect of depressing Dem effort when polls show Rs up, I wouldn't be thinking about any of this now.

One poster here suggested that one of the important things pro K/E bloggers (K/E supporters in general also?) can do is help keep Dem spirits at least as upbeat as they should be based on a reasonably accurate read on how our side is doing.

If this is so, then any impact K/E supporters can have prior to the election on reducing the extent of unsupportably pro-Bush poll reporting may make our job that much easier--and the prospects for a win in November that much brighter.

It should be obvious that the grounds for objecting to unsupportable coverage of polls by media outlets are that this is simply bad journalism. There isn't any reason why legit media outlets should give a rip if it depresses the efforts of partisans on either side.

Look at their history(meaning the media). They did not do their homework on any of the messes that junior and the boys have put us into. Does anyone really believe they will ask hard questions now?Do any of you not think that these thugs did not have this planned after the rethug convention?????? And it is having the effect they wanted. Hell, they are laughting all the way to their dark dens.

Adrock, you wrote:

Thank you for this analysis. My dad keeps on calling me telling me that pollsters ought to be dragged out and arrested. I just sent him this article to calm him down!

Posted by Adrock at September 14, 2004 03:15 PM

Your dad is right. As a group, the polls generally leaned right in 2000 on the Bush v. Gore race. This year they are worse.

When the outrageously conformed polls are added to the mix, they create a perception of a Bush 5 point lead. In fact, it's about tied.

Who is doing the counting?

Who is paying them?

Who are they counting?

Who are they leaving out?

These four questions have to be answered to understand any poll.

Another factor that is being overlooked is the Nader factor. The 2000 Naderites are going to vote again, and they are not going to vote for Nader in battleground states. Bank on it. These are people who take their vote as seriously as any citizen, and many of them are currently registered Greens or independents. A Bush second term scares the shit out of them (us).

There is no way Minnesota, Wisconsin, Oregon, or Washington will vote for Bush. Most of the 5% to 10% who voted for Nader will vote for Kerry. It still may not be enough, but those states will be BLUE. Florida could be affected by this as well, and the minorities in Florida are pissed. If Jeb's cronies don't pull another fraudulent stunt (like purging minority voters and intimidating them at the polls), Florida should be BLUE as well.

I don't want to be rude in pointing out how silly this analysis is, so I will just use Ruy's own words back in June, when he took exactly the opposite position in commenting on an LA Times poll that showed Kerry up by 7% in a sample of 38% democrat and 25% republican. Thus, in Ruy's biased world, it is fine if a poll "over samples" democrats, but not if it over samples republicans.

Ruy's earlier "analysis:"

There are ample grounds for thinking there is, in fact, a surge toward the Democrats and their positions and away from the Republicans and their positions among the broad electorate. A growing Democratic party ID advantage is a logical consequence of that surge, since party ID does not remain stable as political conditions change....Conclusion: there is no good reason to ignore the results of this poll (unless you're Matthew Dowd, of course, who has his own reasons for doing so).