Rasmussen Battleground Update for 10/4

[louis c]

Rasmussen 3 day daily tracking numbers show a more than 1 point Kerry gain from yesterday, with Bush leading 48.6% to Kerry 46.1%.

http://www.rasmussenreports.com

The 5 key battleground states have some mixed and curious results. Remember, these are 7 day rolling averages, so more than half the data is still pre-debate.

Florida, Bush 51%-47% (This is a 2 point Kerry gain from yesterday, and the fifth day in a row Bush has led.)

Michigan, Tie 46% (This represents identical numbers for the fifth day in a row)

Minnesota, Kerry 48%-46% (This is unchanged from yesterday and is 6 days in a row that Kerry has led here by either 1 or 2 points)

Ohio, Bush 48%-45% (This is the curious result to me. Bush did not gain, but Kerry dropped 2 points from yesterday, which is a Bush net of 2 points)

Pennsylvania, Bush 48%-46% (This is also a strange result and represents a 2 point Bush gain from yesterday. All I can think of is that large Kerry gains 7 and 8 days ago have left the poll, but one would think the huge debate bounce would kick in)

My editorial comment:

Don't blame me for these numbers. After all, you get them for free here at DU, I have to pay for the bullshit. I don't believe them, either. Perhaps we're too impatient waiting for the post-debate bounce, but we should have seen something here by now, it's everywhere else.

Well, tomorrow's another day, so stay tuned.

[NewYorkerfromMass]

1. Michigan is solid Kerry. ACT has pulled staff and reassigned

them to other battlegrounds. I can't say I have a lot of faith in these particular polls.

[deckerd]

31. Unlike Gallup they're sauced so as to always show a close race..always n/t

[KoKo]

2. It's all very confusing. On the Ground here in NC in my city I see Kerry/

Edwards yard signs and car bumpers 2 to 1 to Bush. And...I'm in a Repug..though not "evangelical" area. The absence of Bush/Cheney everywhere like it was when we invaded Iraq is interesting. At that time we had cars with two Flags flying on either side and much evidence of Patriotic Zeal and support for the Chimp. One family even had a whole yard filled with flags and Bush signs after Iraq Invasion. Since then ...all has disappeared.

I still see a few "red, white and blue" bows on mailboxes but they may have kids serving in the Military..and I don't begrudge them for that, but all in all, I think NC is in play to turn to Kerry/Edwards. And, we've been told by our Dem Ops that my county could carry the whole state for Kerry...forgetting the Bush base out in the NC hinterlands.

:shrug: Either they aren't able to count the newly registered voters (huge NC numbers of newly declared Dems) and really don't want to skew the polls figuring who can know if those new "registers" will actually show up on election day...is the question.

If I was running a poll I would have to discount the "newly registered" figuring how could one know what they will really do?

They are the "Wild Card" for the Pollsters.

[deckerd]

32. "Hinterlands?"

Voters there count just as much -- literally. There's no electoral college at the county level. For that matter, why is Kerry seemingly giving up Ark and VA?

Does it have anything to do with the 2-1 (in sheer numbers) voter registration advantage that wealthy, right wing Prince William County (VA) has over heavily populated, centrist Fairfax County with its population of federal employees?

What happened to voter registration in the "hinterlands" like Fairfax County, VA? Could that be what is happening in your (presumably heavily populated, centrist, suburban) county as well? If so, perhaps some misinformed individual at ACT or the DNC concluded that YOUR county is also one of the hinterlands that can be safely overlooked because it is insufficiently "vote-rich".

That is a strategy called "relying on base turnout" and we've condemned the GOP rightfully and mercilessly for it.

[KoKo]

37. Yes....the DNC did decide it wasn't worth going after Bush's base but

that we were supposed to focus on getting the Democratic Vote out on Election Day. We weren't even supposed to cavass "Undeclared or Independent" voters.

DNC and my State Dem Organization run by DNC concentrated on areas with higher concentrations of Dems who didn't turn out in the last elections or only vote in Presidential elections.

Hey...I think our campaigns are corrupted by so much polling and demographic info that the Dem voters in Repug areas are supported, enough. But, they have limited resources and that was the call they made. It doesn't mean you can't do something in your area on your own to turn out your fellow Democrats wherever they are. It's hard if your area is very Repug but go on the web and get some info from your County Party about where to pick up signs and find fellow Democrats in your precinct.

Maybe you are already doing that..you didn't say. I'm sorry if you were offended by my use of "hinterlands." I meant areas outside where DNC is concentrating...

[mvd]

3. PA 2 points for Bush?

No way. x(

Rasmussen's starting to get suspicious if you ask me.

[tritsofme]

7. Why?

Because he's showing results that aren't always favorable to Dems?

If Rasmussen has anything going for him its consistency.

After his tremendous fuck-up in 2000 he probably has the most to prove out of any pollster.

[mvd]

8. Because he's going away from trends

PA has been on Kerry's side in most polls lately.

[fertilizeonarbusto]

4. polls are meaningless

the record number of new registered voters will make them insignificant because those people never get polled. I'm in PA and I've yet to meet one person among the dozens registering for the first time that is voting for Shrub.

[mvd]

12. True

I have 3 Repuke relatives who don't like Bush - one is voting for Kerry, one is thinking about voting for Nader but can be persuaded to vote Kerry, and one older one might stay home.

[Teaser]

14. They are not meaningless..

however, they are a very imperfect measure of a very imprecise quantity.

[Bombtrack]

27. this is a moronic sentiment. It isn't backed up by reality and logic

why would the new registered voters not get polled? They don't poll off of registration lists they poll off of random calls where they ask whether people are registered/planning to vote etc.

[PermanentRevolution]

28. It's not moronic

Each pollster has their own methodology for deciding which voters to include as "Likely Voters." Generally speaking, if you've never voted in an election before, you're thrown out of the pool. So all the newly-registered voters are excluded from the "Likely Voter" polls. I'm not sure if these numbers are "Likely" or "Registered," though.

[mahina]

29. You may be right,

but I remember feeling so sure that Dean was going to prevail in Iowa because all the young folks with cell phones weren't represented in the polling.
Whoops.
Of course I love Kerry now, with a very big love.
Still I remember feeling so sure the polls were wrong.

[spooky3]

5. thanks for posting these.

[lancdem]

6. louis, remember state polls are a lagging indicator

and one debate won't change the race overnight, but I expect if Kerry continues his strong performance in the next two debates, he'll be in very good shape.

[deadparrot]

9. These numbers don't take into account the millions of new voters,

more registering everyday, mainly in these battleground states.

Remember folks, the foreign policy side is where Bush is strongest. Many were expecting Kerry to be slaughtered in that first debate. There was talk that * could "put the election away." By winning, Kerry's already surpassed a huge hurdle. Domestic policy is where * is extremely vulnerable. Hopefully, Kerry will not ease up, but go harder.

[tritsofme]

16. Why wouldn't they count new voters?

The main criteria most pollsters use is by asking "Are you currently registered to vote" or some variation thereof.

[deadparrot]

17. Because...

from the link,

"These results are based upon a survey of 3,000 Likely Voters conducted since Thursday night's debate."

Likely voters are defined as voters who have a history of voting in elections, meaning that the 18-22 crowd (first opportunity to vote in a presidential election) aren't being counted, plus the millions of older new voters that are currently flooding voter registration offices. There are a lot of people pissed about *'s policies, and plan to show it. There was an aritcle awhile back about how Democratic registration was up 250% from 2000, while Repug was up 25%.

[tritsofme]

21. As far as I know Rasmussen doesn't make their LV model public

So new voters very well could be counted as likely voters, unless their LV model is known we have no way of knowing how they determine what is a likely voter.

[deadparrot]

22. Perhaps, but

most polling agencies define likely voters as what I have already posted. Unless they say otherwise, I assume that Rasmussen follows the same formula.

[deckerd]

33. That's 250% of a comparatively marginal % of the total vote.

And only a certain percentage of new registrants end up at the polls. Maybe higher or lower than the percentage of all registered voters who end up at the polls, depending on how they got registered -- the entire GOTV model is dependent on the knowledge that most new registered voters have to be essentially (politely of course) browbeaten into doing their civic duty, even if they dislike * they dislike politics in general just as much.

Bottom line is, fluctuations in the preferences of early undecided
voters (likely voters - mostly semi-affluent 30 and 40 somethings)
outnumbers all new registrations combined by a huge amount. These
undecideds generally break for the incumbent in mid-October, or
else they split between not voting and breaking late for the challenger.

The election is still an unresolved popularity contest with the mass of undecideds. Whaddaya think's wrong with this country

[deckerd]

34. Of course, turnout makes a big difference too

But registrations is not a predictor of (overall Democratic) turnout.

That said, absentee balloting is heavy at a time of relative Bush strength. Which is not necessarily good for Kerry unless registrations is indicative of a groundswell of motivated base voters, which is apparent in some states, but not others (in VA and possibly Ark, perhaps Kerry has concluded the Repub base is simply more riled up in those states and Dems insufficiently so. that would explain the discrepancy in numbers with only 13,000 new registered voters in behemoth Fairfax County VA -- compared to 26,000 in conservative, semirural Prince William County.)

[wryter2000]

10. Thanks, Louis

The picture will get brighter.

[Zero Gravitas]

11. Turnout is key

Forget the polls, what will decide this election is turnout, the higher the turnout, the bigger the landslide for President Kerry.

[louis c]

13. kick

[59millionmorons]

15. Thanks for sharing

I dont put much stock in Rasmussen after there 2000 results, but it's nice to get a peak.

Thank You

[Awsi Dooger]

18. I have noticed Rasmussen is always high on Bush's favorability number

It has been above 50% for weeks, usually above 52% and that is higher than virtually any other source. Down to 52.1% today while many others are below 50% now.

I would appreciate it if the approval rating were also posted in this daily thread, since an incumbent's approval rating has been demonstrated as a critical factor. Thanks.

[louis c]

24. It's the way the question is asked

"Do you have a favorable or unfavorable opinion of President Bush?"

A lot of people might like him, but think he sucks as a president.

When the question is asked correctly (Rasmussen doesn't) it should be "Do you approve of the way George Bush in handling his job as President?", it shaves at least 6 points off his total.

[Awsi Dooger]

25. Thanks, that does make all the difference

I'll ignore their favorability number from now on

[shockingelk]

19. Thanks

3 points down is nothing. Good turnout can overtake 3 points easily.

Gore was 5 points down nationally in 2000 al la Gallup.

How accurate have state level Rasmussen polls been?

[Mike L]

20. I have given up on Rasmussen. His numbers move in funky ways, none of it

logical. You would think there would be some consistency to the numbers with the 7 day rolling totals. I haven't found it.

[sonicx]

23. He has odd polls, but i'm still not worried about PA

or even the other close swing states and the national race. We have bigger voter reg and we'll have bigger GOTV.

[fujiyama]

26. This is just plain strange

I really don't think it makes much sense...Sorry, MI is NOT tied. Bush is also NOT in the lead in PA...and while he may be up in FL it's not by 4-5 points (maybe 2).

The only two that make real sense are OH and MN.

[mahina]

30. Oh and thank you-

I appreciate the info!

[VOX]

35. Thanks, louis -- always interesting...

even is a bit disappointing and/or puzzling sometimes. These numbers will be seesawing around over the next couple of weeks.

Your efforts are appreciated! :toast:

[louis c]

36. You're always welcome

[Spiffarino]

38. The usual "Likely voter" poll

The fine print:
This telephone survey of 781 Likely Voters was conducted by Rasmussen Reports from Sept 27-Oct 3, 2004. The margin of sampling error is +/- 4 percentage points with a 95% margin of sampling error.