« New Voters for Kerry | Main | Analysis of Trends in Major Media Polls »

Bush's Battleground Blues (Continued)

A few days ago, I highlighted some recent polls that showed solid leads for Kerry in the battleground states as a whole, states that were split about evenly between Gore and Bush four years ago.

Since then, Democracy Corps has released new data showing more of the same (a 7 point lead for Kerry in the battleground states). And Mystery Pollster looks at a substantially wider range of recent polls and finds Kerry's battleground performance running ahead of his national performance in every single one. As Chris Bowers points out over at MyDD, these data show Kerry averaging a 49-45 advantage in the battleground.

And, not to pile on, but check but the latest unemployment data from the battleground states. Not a pretty picture, by and large, for BC04: Wisconsin and Iowa show increases in their unemployment rates in the last month and Ohio's remains stubbornly high at 6 percent.


I would like to know what bloggers and the like can do the make explosives-gate as important a political issue as possible.

Explosives-gate is, of course, the recent reports that over 350 tons of high explosives were taken from an unsecured ammunition facility and that the ***Bush administration attempted to keep news about this from being released*** until after the election.

I think this should really dent Bush's perception that he can be trusted to wage either the war in Iraq or the war on terror.

Quite possibly, something very real might lie behind the disparity between the head-to-heads in the battleground states and the head-to-heads nationally.

It may be, for example, that the message that gets out to voters in the battleground states is qualitatively (and perhaps quantitatively) different from what gets through to voters in other states. One obvious difference is that the battleground states get more information, because there are so many campaign events and so much advertising, while in the other states it is only the free media that communicates anything at all. Moreover, the message of the candidates is also mainly geared to the circumstances and concerns of the battleground states.

It's quite possible that the Kerry message that gets through to voters in battleground states is just distinctly more effective than the message that the free media communicates in the other states.

If true, one very curious consequence might very well be that there will a large gap between the popular voter and the electoral vote, with Bush doing very well, perhaps even winning, the popular vote, by running up the vote in non-battleground states (in both the states he will certainly lose and the states he will certainly win), but losing decisively nonetheless in the electoral vote.

Zogby has Bush turning it around in OH....Any info on the Zogby detail -- party ID, etc.?


Ruy, you say, "And, not to pile on, but..."

As far as I'm concerned, on this subject you can pile on all you'd like. And more.


I've been checking out this site lately and they have gone from Kerry in the lead to swinging wide for Bush with the recent polls. As he says himself, he isn't sure he believes most of them.

Interestingly, he is taking undecided voters into account; saying that they usually go for the challenger.


Ruy, you should do a post (or have a link section) with all of the poll data sites you are aware of (or, the good ones anyway.) Its nice to see how different people read the tea leaves.

Ruy please comment:

Zogby Oct 25:

“The President has opened up a 12-point lead among Independents and now also leads among those voters with active passports. While each candidate is polling solidly among most of his base constituency, Kerry should be concerned that his numbers among Hispanics appear to be anemic. We are finding the same in states like New Mexico, but we are currently engaged in a large national poll of Hispanics—and will be able to
tell shortly if the findings in this tracking poll correspond to a sample of 1000 Hispanic voters"

I keep seeing references to campaign "internal polls". How can these really be different from the published ones? Aren't they taken by the same doofusses, and subject to the same rather larger errors? Don't organizations like Zogby also do internal or private polling for politicians? I wish someone could shed some light on this.

I certainly empathize with the unemployed in Ohio, and understand that the official unemployment rate is lower than the real unemployment rate; but I wonder how they would like to try on the unemployment rate here in central California--it is typically double the current rate in Ohio *when times are good.* When times are bad, it goes over 20%. And the area votes Republican (big surprise).

He sure does have a link section, scroll down, it's on the right.

Electoral-Vote.com is a good site. Which is why after visiting it this morning and seeing Bush jump ahead signifincantly in their prediction, I had to come here to make sense of it.

There is one thing that bothers me though. Sure, we keep on saying that the undecideds generally go to the challenger this close to the election. Also, we've been complaining about the LV qualifications and questions in that they underrepresent minorites and youth. But couldn't LV question underrepresent unintellectual Bush voters as well? Say they only get 3 out of the 7 questions wrong and so the pollster doesn't consider them likely, that doesn't mean they AREN'T voting for Bush. Both sides have quite an energized base. And I think we've almost reached the point that undecideds really aren't going to be deciding this election, "base turnout" will.

Is it me or has Zogby's polls gone crazy in the last week?Bush's lead is widening, Kerry's support is dropping out among all groups. He has this quote today

"The President has opened up a 12-point lead among Independents and now also leads among those voters with active passports."

Virtually every other poll I've seen has Kerry ahead by 10 points or so among independent voters. What does the active passports comment have to do with anything, anyway? I have no idea what influence voters with passports have.

To make things more strange, Kerry's down in almost all the swing states. The ones that seem most strange to me are the 4 point lead in Colorado (Kerry 49, Bush 45), and Bush's 5 point lead in Ohio (Kerry 42, Bush 47).

I had been believing Zogby's numbers were more reliable than most, as he and CBS were the only ones to predict the 2000 election for Gore. The also always had the race closer than the rest of the polls. These numbers just keep getting more bizarre though.

Any thoughts?


The site (www.electoral-vote.com) does swing widely. They use a number of polls that are paid by political parties (notably, the Strategic Vision poll paid for by Republicans). These tend to favor the people paying for the polls.

Zogby's battleground state polls out today seem to be throwing everything towards Republicans. However, the numbers are out of whack with conventional wisdom (undecideds going toward Bush, Hispanics towards Bush, etc.?). I am not ready to accept those numbers. It's almost like they got something backwards (????). I know based on the phone calls I made in San Diego County for the Kerry/Edwards campaign, I did not have a single Hispanic out of 30 that said they favored Bush. I think we need to see a few more days worth of data before we believe those numbers.

Whoa! Rasmussen today shows a 2-point lead for Kerry! This seems to run counter to the Zogby poll today, which shows Bush's lead expanding to 3 points...but I think GW had an unusually good day of Zogby polling on Friday or Saturday.

Alan is right, these tracking polls are nuts. Rasmussen has Kerry two points ahead, Zogby has Bush ahead +3, and TIPP has Bush up by +8! The ABC and WaPost are not out yet but yesterday they had Bush a mere +1.

I always believed the tracking polls provided a better snapshop of public opinion but it seems the survey polls are more consistent lately.

It will be fascinating to see how the news about the 380 tons of missing pure high explosives impacts the final days of the race.


Kerry leads Bush by 2 in the latest Rasmussen. Anyone have any idea why there is a disconnect between Rasmussen and Zogby? And, yes I remain obsessed!


The trends in the battleground states keep being very encouraging. The national polls keep having Bush in the lead, though. Now, there are the legitimate concerns we have about all those Gallup-esque likely voter patterns. But isn't the possibility getting more real by the day that this time around Bush might win the populare vote, however close – and lose the electoral vote?

It would be somewhat disappointing for Kerry and his presidency. But, in a way, it would be biblical justice, wouldn't it...?

Are the polls accurate? Is Bush really up by 2-3 points? Zogby has Bush up by 2, for instance. I went over to 2.004k.com and counted up sure state winners for Kerry. I came up with 253 electoral votes without even trying. That leaves Florida, Iowa (my state, where early voting gives Kerry the lead, and GOTV in heavily Democratic Johnson County alone is going on 24/7), Colorado, Hawaii, and Arkansas...what am I forgetting?--all states Kerry could easily win. Anyone?

This is a good blog, but annoyingly one-sided at times. I always need to hold my nose and browse the 'winger blogs for the polls and news Ruy does *not* mention. It seems this week's Monday hasn't been a good day for Kerry in the state polls; Slate, electoral-vote.com, PollingReport.com and (of course-) the right-wing pair of FederalReview.com and ElectionProjection.com all report "Shrub" is ahead, albeit not by much. But thanks to "Shrub's" gains in the smaller Mid-Western states and the Southwest (including Hawaii of all places!), there are now reasonably credible scenarios where Kerry wins *all* three major battleground states (PA, OH, FL) and still loses. I don't like this at all.
I don't think this signals the beginning of a trend (last week was generally pretty good) since the poll movements have been fairly small. But I will be nervously watching the President's approval ratings during this week. If they suddenly start moving above the 50% mark, it's not a very good sign for Nov.2.


What you have to remember about electoral-vote.com is that he only uses the most recent
poll as opposed to doing any averaging or tracking.
Also, he doesn't do things which I do as a
matter of course: 1) subtract 2 points from
Bush on all SUSA, Rasmussen, Strategic Vision,
and Mason-Dixon polls and give them to Kerry;
2) Throw out Gallup polls altogether; 3) simulate new-for-this-election voters by giving a net 2% to Kerry where applicable; and 4) give
undecideds to Kerry by a large margin
(67%/75%/80%/86%, to see the effects of each).

from electoral-vote.com website: highlighting wild swings in current polling.

In contrast to previous Mondays, there are many new polls today, with 19 states getting new numbers (although most didn't change sides). In addition to the ususual polls, Zogby has begun daily tracking polls in 10 battleground states, which I will also toss into the hopper. According to Zogby's polls, conducted Oct. 21-24, Bush is currently leading in six states (FL, NM, NV, WI, IA, and OH), while Kerry is leading in four states (CO, MN, PA, MI). Some of these results are very surprising. Is Kerry really leading by 4% in Colorado? Is Bush really leading by 5% in New Mexico? I don't believe either of those. They are in conflict with too many other polls. Another example: the current Ohio University poll gives Kerry a 6% lead in that state, whereas Zogby puts Bush ahead by 5%. The MoE on these polls is 4%, so an 11% change in a couple of days in a state with so few undecideds is impossible. I think there are serious problems with the all the polls.

Re: zogby

why do people keep insisting on treating one pollster as if they are the end all- be all of polling on either the left or right? There are multiple polls so why not take the approach of reading ALL polls in context. Now, as one poster here mentions below, the problem comes when its hard to figure out the context.

There is encouraging news Monday from two tracking polls that have had Bush ahead for weeks and weeks until today. Kerry now leads Bush in the just updated Washington Post tracking poll and the Rasmussen tracking poll. No doubt these polls have their flaws, but as relative measures (relative to themsleves) they are both showing a clear tend over the last three days of Kerry picking up strenghth. I belive it's starting to break pretty clearly for Kerry.

www.electoral-vote.com is an ok site, but very simplistic. they take the latest poll and go with that, no detailed statistical analysis. the Hawaii issue is a good one, Hawaii is NOT going to go to Bush, every poll has kerry with a large lead and anybody who have leved there know its as likley to to to Bush as MA. the problem was a bogus poll reported in some blogs had bush barely ahead. Everbody is discounting the poll, yet www.electoral-vote.com used it to give state to bush

If you really want a true analyssi of the state of the college and the cahnges for Kerry go to

Sam Wang does what is know as a meta-analysis. it may be beyond most readers math level, but he does several million calcualtions and calculates the probability of every possible outcome and then does a 50th-percentile (expected) outcome, as well as a 95-percent confidence interval.

for today its

Predicted median with undecideds: Kerry 307 EV, Bush 231 EV (probability map)

Median outcome, decided voters only: Kerry 259 EV, Bush 279 EV

Among decided voters: Bush leads Kerry by 0.5%

so the election is very close with just decideds, taking into account undecideds Kerry is way ahead.

I'm encouraged by this post, but there are caveats. I don't see that the link identifies the "Battleground States". Also, it identifies a 7% lead based on Democracy Corps polls, which are partisan and lean democratic, at least by 2-3 points, typically, compared to media polls. Also, Kerry needs to win a lot more of the battleground states than Bush, since the Bush safe states comprise a lot more electoral votes than the Kerry safe states, and most battleground states are blue. In short, this is a very "tweakable" statistic, and we need a lot more information.

AS to kerry needing to win more states:

with regard to red blue states that are strong for either one diff of just 20 votes:
Bush has 216 EV
Kerry: 194 EV

Weak states, where the state is at some risk to go to one or the other, usually trending but within margin of error, and not taking into account undecidededs

Bush Weak: 74
kerry Weak: 45

Bush is at more risk (note the 30 vote difference)

Undecided EV where they are tied or different polls og to each one: 128 EV

Taking in account just decided voters the election is very close, if undecideds are factored in Kerry is ahead