My daughter had to read a newspaper and watch the local news as part of her homework assignment today. I had only seen about three George W. Bush ads since the last debate, so I was surprised to hear a sudden flurry of his ads:

News story: A cargo plane en route from Tokyo to New York was forced to make an emergency landing in Detroit when it lost an engine after refueling in Chicago. Officials said, that as no one reported spotting the engine or being harmed by it, no effort would be made to search for the lost engine.


The state now has the third highest unemployment rate in the country…


A man … is charged with possession and distribution of child pornography. So far, more than 1,000 pornographic pictures, some featuring young girls, have been found on the suspect's computer and several CDs.

Yes, George approved that message too…

Eventually we got to the reason for the upsurge in the ads. A Detroit News poll released today showed:


Another poll result in the same article showed Michigan voters listing jobs and the economy as the most important problem facing the state had risen from 43% to 59% since January.

For the moment, I am going to take these results of a poll taken by a Republican leaning newspaper with a grain of salt. My hope is that since W's gain in the polls was made while his advertising was cut, that if he continues to advertise heavily, he will raise Kerry's share again.

Pursuant to a conversation earlier in the day, I took my daughter to a BBC website, where links to the international news brought me to a close to home article, Treat polls cautiously, experts warn.


The article discusses several variables in the phrase “margin of error” to support a question asked early in the article,


In the past few days, I have learned that:

1. Item 1 Polls of “likely voters” are excluding anyone who was not old enough to vote four years ago.
2. Item 2 Cell phones cannot be called by pollsters.
3. Item 3 Registration among college students is at an all time high.
4. Item 4 Due to caller ID, getting someone to answer the phone is more difficult than in the past.
5. Item 5 The amount spent on advertising in any state is now being shifted day to day depending on current poll results. (If we’ve already won/lost the state, why waste money?)

This year apparently, “Can we trust these polls?” is becoming a very interesting question.

To discuss:

Based on conversations you have with friends and casual acquaintances, do the published poll results seem to reflect any reality of how you actually expect the election to turn out where you live?

Does the level of political advertising where you live change dramatically with every newly published poll?”

1. Well, my friends seem to believe the stories that this will be a very close race, maybe even closer than 2000. I'm not so sure, however. I'll explain in a moment.

2. I live in Illinois, a solidly Democratic state. There have been a few Bush ads, and even fewer Kerry ads where I live. Mostly I think, due to the fact that everyone knows it's a heavily Democratic state. Why bother to heavily advertise if you already have the state sewn up, or if there is no chance of converting enough voters to make a difference?

I'd like to go back and address your list, though Curmudgeon.

Item Five, I believe I covered in my second point of reply.

Item One, actually makes some sense. Traditionally, a lot of people that register, fail to vote for one reason or another. So "likely voters are those who have voted in at least one past election in the last four years, and better yet, also voted in the primaries.

Item Four I can understand, but there are still people who either don't have caller ID, or who don't mind taking part in polls and surveys. It may be harder to reach some folks for polling now, but it's still doable.

Items Two and Three are the ones that really have me wondering how accurate the polls are. Your right, Curmudgeon, cell phones can't be called by pollsters.

And college age kids (18 to 24) are registering at an all time high. More are registering now, than at any time since the vote was given to 18 year olds. And since many have not voted before, or even registered before, they will not be counted as "likely" voters, even if they are contacted. But that's the biggest problem.

I read a few weeks back that some 35% of college age kids have only a cell phone as their primary phone. 35% have no land line for pollsters to call. And as you noted earlier, those with land lines, and answering machines or caller ID's have the ability to screen out calls like polling organizations. I would imagine that students would be far less tolerant of those calls than would, say, people of our generation.

I think that the voting may not be as close as some imagine, and it may well be due to a new generation of student voters. Voters that the pollsters have no real way of accurately gauging.

No, we can't trust these polls. I think that Bush would need a 7-8 point lead in these polls just to be even with Kerry in reality. The number of people who will vote in this election is being seriously underestimated, and the pollsters are not hitting the new voters. Kerry will win by a wide margin on Nov 2, and one of the big stories will be... "How could all of the polls have been so wrong."

Gallup has been skewing its polls toward Bush by sampling 5% - 8% more Republicans than Democrats, despite all indications that this sampling is unjustified. http://www.theleftcoaster.com/archives/002806.html

CNN has been highlighting these Gallup polls and others favorable to Bush, while ignoring contemporaneous polls that favor Kerry. http://www.mediamatters.org

They're trying to inflate Bush's numbers because swing voters can be heavily infuenced by the desire to have voted on the side that ends up winning. Distorted poll results can sway enough votes to decide an election, and they also discourage Kerry's volunteers. As long as Gallup and their broadcast outlets like CNN retain credibility, we're at the mercy of the sick fiction they pass off as news. http://www.carefulvoter.org

It's interesting that you take this view, since Kerry is ahead in virtually NONE of the polls, regardless of polling organization. As a matter of fact, one of the only polls that Kerry was ahead (nationally) in a 3 way poll was a Gallup poll RealClearPolitics Poll Average. The same is true for head-to-head polling Head-to-Head Race: Bush/Cheney vs Kerry/Edwards

Perhaps you know of some surveys that are not included in the RealClearPolitics average of surveys?

Based on conversations you have with friends and casual acquaintances, do the published poll results seem to reflect any reality of how you actually expect the election to turn out where you live?
Actually, I am surprised by some of the people who are supporting Bush. In an informal poll in my church Sunday night basketball league, Bush is leading 7-1 among people who I would have guessed to be Democrats. The only Kerry guy seemed more enthusiastic (he volunteers for Kerry), but the others were happy with 2 things: tax cuts and Bush as commander in chief.

I live in Ohio, by the way, in Cuyahoga County, the most Democratic county in the state.

Does the level of political advertising where you live change dramatically with every newly published poll?

It doesn't seem to. I dunno. I am surprised at how many Kerry ads are broadcast during the Rush Limbaugh Show and on Fox News, however. If I had to guess, I'd say the ads are 50-50 Bush-Kerry. This is in Ohio, perhaps the most important swing state.

1. The published polls seem to be a reflection of public opinion. I live in Dallas, TX where an overwhelming majority of voters are Bush people and republicans to boot. However, my occupation dicates regular travel as does my status as a Marine reservist. I was in Little Rock one day last week and, as we'd suspect, there was a large dem constituency. At the same time, I spent 2 days in Memphis this month where the vote was markedly split. With that in mind, it seems that the electoral predictions are accurate. (just a gut level opinion)

**In reference to the statements about the gallup polls and leaning towards the republican candidate, please consider the fact that ALL 3 polls show a Bush margin. Is this because they all are biased??? The polls all show a margin of error. This is the nature of polling. While many people remark that with the margin of error that Kerry and Bush are even, you have to consider the fact that margins of error are +/-, and could very well fall towards GW. **

(Also consider the fact that even the proposed electoral votes show an edge for Bush as well.)

2. In Texas, as I suspect with other "red" states, advertising probably doesn't change much after polls are published. I would suggest that in swing states that advertising is purchased in large volumes and that for the next couple of weeks we will be barraged by political ads. I don't have any objective information about advertising changes after polls, but it would be perfectly sensible.

In terms of student voters, I would agree that many do not possess land lines. I also have two SMU interns in my office and believe that many young people (not that I'm all that old!) are politically charged. However, in the south, there is a large republican constituency among collegiate voters. At the same time I sincerely wonder how many young people actually are registered to vote and how many will take the time to get to the polls? Right out of college I worked in Shreveport, LA and employed numerous college juniors and seniors (as I do now). Often times they have very "large ideas" coupled with very little action. We'll see how that all turns out...

I can't wait to see what the demographics are of the people that actually get to the voting booth. Consider the fact that the elderly and minorities are far less likely to vote, which would often times vote for the democratic candidate. Just a thought.

How many eighteen years old, recently registered voters, have they called when doing this polling? Just about everyone at UCSC is registered to vote, so if there leaving out recently eligible voters thats could cripple the accuracy. Trust me, we're going to show up in swarms. At least, more so than any other election.

But the national poll, at least point, means very little. Kerrys in a good position because Penn. is just about in the bag, putting it on Bush to win both Ohio and Florida. We'll see what happens though.

I live in California. No one campaigns in california, except Nader.

Don't think the polls are reflective of reality this swing around. Much has changed over the past three or so years. The 2004 election will not resemble the 2000 election, from what I'm seeing.

Of the many post-election talking points, I think the swing from apathy to intense involvement will be noted. Additionally, the fundamentals of polling will be in question as the habits of citizens change with the changing technical landscape, as pointed out with the cell phone example. I might add that many have opted for pay-as-you-use cells rather than the monthly fee plans.

I have one of these, and I only turn it on for making outgoing calls.

Another wrinkle that will be explored is the tightening of the middle class budget. Consider an unemployed person using the services of the unemployment office (phone, Internet) rather than paying for a land line and ISP. Other unemployed folks might be using a friend's land line for contacts.



Oh, it is SO nice not to have cable TV or rabbit ear broadcasts! I am completely unaware of political ads, except those shown for fun on the Internet.

I thought I'd comment briefly on this because Zogby has some data from a cell phone poll.


I'd say that we can't trust traditional polls anymore because they are behind the times. The electorate is changing, especially this year, and pollsters in general have yet to react to that. I think Zogby's poll is one of the first that actually takes into account cell phone users. The fact that millions upon millions of people will vote in American Idol via SMS text messages should tell pollsters something about how times are changing.

At my home, I have a phone but no one I know has the number. We use the number for creditors, etc and it is connected to the call box downstairs. Most of the time I don't even answer it, let the answering machine catch it and instead use my cell phone for everything. I'm 26 and this is not unique for my generation.

1.The people where I live are mostly outside of the conventional wisdom of the polls. I live in a die hard republican area, so I thought that Clinton would lose in '96-it was just the geographical area around me that wasn't with it and tends to be out of it even when Bubba's re-election became fairly obvious.

2.)With only five electoral votes, we don't see any. We do get radio ads on Omaha stations due to what goes on in Iowa.

Thanks Cube Jockey, for your remarks on the Cell Phone polling.

My day started out with buying a copy of the Grand Rapids Press for a closer look at the front page grapics. I had grown up in Grand Rapids and worked in Midland County for 30 years. Ottawa County, a few miles South of here, I noticed had voted Republican in at least the past 11 Presidential elections. There is a good chance, I have heard, that it will go for Kerry.

As I was handing out candy tonight, I was met with something I had never encountered before. The parentss of very young children were saying, "I haven't had time to volunteer, but this seemed a good opportunity to ask my neighbors to vote for Kerry on Tuesday." I got similar, but not identical messages from 15 - 20 sets of parents. Even with the signs removed from the yard for the night, no one asked me to vote for Bush. I can't recall that I had ever been approached by trick or treaters asking for my vote before. It seemed to be a very spontaneous event.

As the graphics in the newspaper so eloquently illustrated, this area normally votes Republican.

Zogby was on the Daily Show- and when Jon Stewart asked him who he thought would win- he didn't even hesitate- "John Kerry"- when asked if would lose sleep over it if he was wrong, he replied "yes, I am paid to be right, this is my livelihood, it agonizes me when I am wrong"- and then explained "When an incumbent has it this close, he is in terrible trouble, and I think the young vote will sway it heavily Kerry".

Yeah, CR, he was on "Money Talk" with Bob Brinker over the weekend, and pretty much said the same thing.

The point I found interesting, though, was that Zogby said that polls asking if the country was going in the right direction or wrong direction, clearly were showing that the electorate thought it was headed in the wrong direction, by an average of about 8% points. If that's the case, I don't see how it can possibly still be this close.

I don't think it is that close, and that will be the failure of all of these polls, but I guess we won't know for sure until Wednesday. For now, the only thing in the realm of fact I have to base my opinion on is the following:



To sort of sum up, the factors I honestly believe will throw all the pollsters for a loop this year are:
- Record voter registration numbers
- "Importance" of the election and the effect on turnout
- Behind the times with technology and trends - in other words all of the things Curmudgeon mentioned such as not polling cell phones, etc.
- Defections from the other side due to Bush breaking with his own party on several issues.

Zogby called this race for Kerry months ago. Not one month ago, but almost SIX months ago!!



I am not saying he is biased, but I am saying that he has a vested interest in helping Kerry pull this out. His numbers at that point were more favorable for Kerry, by about 6 percentage points. He has not re-evaluated his position, interestingly.

His quote "I am paid to be right..." and yet he called this election 6 months before anyone stepped into a voting booth. This in the quote unquote most divided election in history.

It boggles the mind that a pollster, instead of reporting what people are telling him, is making predictions.

As I have said earlier in this thread, there is also a groundswell for Bush that isn't being considered. Record voter registration does not mean that only Democrats are getting out the vote. Keep in mind the millions of Christian Coalition types that did not turn out for Bush because he wasn't religious enough...

This election will swing on three states: Ohio, Pennsylvania, and Iowa.

I do not think the race will be razor thin, as the 2000 election was.

I think you nailed this one right on the head. With more people using cell phones, more people with caller I.D., I certainly have lost a lot of confidence in the polls.

I no longer have caller I.D. since they started the Do Not Call lists, but when I did I would not have answered a phone call from a polster. I wonder how you factor that into the equation?

If there will be a real loser in the election, I hope it's the polling companies.

Edited to add:

I wonder if I will ever live to see the day when I hear a "talking head" say, "The Candidates are both heading to Nebraska to try to capture it's all important 5 electorial votes".

Well, The Red Sox DID when the World Series! I can dream!

Zogby vs. reality. The man is a complete hack. And I don't think it's cell phones neither.