3 Weeks Out: The Polls And LV Screens

By Big Tent Democrat, Section Elections 2008
Posted on Mon Oct 13, 2008 at 08:35:43 AM EST
Tags: (all tags)

Share This: Digg!

With 3 weeks to go, the polls have all switched to a LV (likely voter) model. It is important to understand the each organization has its own likely voter model. Gallup invented its famous one (the 5 questions) and, as far as I know, has not varied it much. As a result, Gallup LV polling has been subject to wild fluctuations. By contrast, Rasmussen (and others, R2000 for one, I think) on the other hand, has switched to a party ID likely voter model and is more stable. And of course each polling outfit has its own general house biases and flaws. (One misconception to be addressed here is about the Bradley Effect. That turns up, for the most degree and if at all, in exit polls, not in pre-election polling.) Later I will round up some old posts from 2004 on the LV model controversies. But for now, let's look at the polling 3 weeks out.

DKos/R2000 has Obama up 12, 52-40. WaPo/ABC has Obama up 10, 53-43. Gallup's new LV polling has Obama up 6, 51-45 (or 4, 50-46, using their 2004 model.) Hotline has Obama up 8, 49-41. Battleground has Obama up 8, 51-43. Ras has Obama up 5, 50-45.

More . . .

The final big event of the campaign is Wednesday night's Presidential debate at Hofstra University in Long Island. After that, it is a sprint to the finish.

By Big Tent Democrat, speaking for me only

< Krugman Wins Nobel Prize For Economics | The Politics of Contrast Triumphant >

Legal

Nothing on this site should be construed as legal advice. TalkLeft does not give legal advice.

TalkLeft is not responsible for and often disagrees with material posted in the comments section. Read at your own risk.

3 Weeks Out: The Polls And LV Screens | 71 comments (71 topical)

Readers Must Login or Create an Account to Comment

I think likely voter models are fine (5.00 / 0) (#1)
by andgarden on Mon Oct 13, 2008 at 08:43:02 AM EST

for low turnout special elections. But for a Presidential, election, they just introduce noise.

What kind of person is who willing to take a poll about how they're going to vote can be determined an "unlikely" voter?

Well (5.00 / 1) (#3)
by Big Tent Democrat on Mon Oct 13, 2008 at 08:46:46 AM EST

My idea is to publish both RVs and LVs , especially on Election day and test the LV models. The pollsters never do that - and for good reasojn, who would need them if their most important model is not worth the pape rit is written on?


: Fair enough (none / 0) (#6)
by andgarden on Mon Oct 13, 2008 at 08:50:20 AM EST

You betray a ... (5.00 / 0) (#40)
by Robot Porter on Mon Oct 13, 2008 at 10:59:56 AM EST

rather simplistic knowledge of human nature.

A lot of people who answer phone surveys are simply bored. Nothing interesting on TV. The spouse isn't paying attention to them. So they'll talk to a pollster.

The LV model can help screen for this.


The question is: (none / 0) (#43)
by andgarden on Mon Oct 13, 2008 at 11:04:51 AM EST

why trust a model more than what people say they're going to do?

I agree with you that a likely voter model should be better than RV, but how do we know that the model in any given election actually improves the data?

It seems like a guess on top of a guess.


Because that's not how it works ... (5.00 / 1) (#51)
by Robot Porter on Mon Oct 13, 2008 at 11:18:34 AM EST

they don't just use a likert question to determine likely voters.

They use what's called "a likely voter index" which is a series of questions which determines their likeliness to vote.

They've been found to be about 75% accurate, which isn't perfect, but it's not bad for an index-based assessment.

It's certainly better than a guess.


I'm sure they have a very (none / 0) (#52)
by andgarden on Mon Oct 13, 2008 at 11:22:25 AM EST

sophisticated model. But I am always suspicious that a model that applied to the past makes any sense today. Maybe they're right--who knows?


I wouldn't be overly ... (none / 0) (#54)
by Robot Porter on Mon Oct 13, 2008 at 11:37:52 AM EST

swayed by talk of the past vs. the present if I were you.

The basic rules of politics remain unchanged.

The party in power screws up. The party out of power wins.

That's as old as elections themselves.


Well sure (none / 0) (#56)
by andgarden on Mon Oct 13, 2008 at 11:45:06 AM EST

But in that case, why even bother to poll? Let's plan the party, right?


: I don't pay a lot of attention to ... (none / 0) (#59)
by Robot Porter on Mon Oct 13, 2008 at 12:18:27 PM EST

polls. I look at them for trends.
But I pay more attention to demographics and historical voting patterns.
And my predictive abilities are as good as poll quants.

i'd be curious to know (5.00 / 1) (#13)
by cpinva on Mon Oct 13, 2008 at 09:03:49 AM EST

if there is historical data available showing the difference between the number of new registrations, vs those that actually voted.

it's my understanding that a huge chunk of the NRV's are part of the group that historically has the lowest level of actual turnout, on election day: 18-21 year-olds.

should that be the case, then the NRV's don't really matter, in real terms.

: I would be more curious to know (none / 0) (#70)
by patriotgames on Mon Oct 13, 2008 at 03:55:51 PM EST

how many people in the polls will NEVER vote for Obama, but out of guilt, or whathave you, tell the pollster that they are voting for him.
We saw that a lot in my state for Governor a few years back. According to the polls most people said they really liked one candidate, but the election results were a landslide for the other candidate.

Polling is great (none / 0) (#2)
by Fabian on Mon Oct 13, 2008 at 08:43:49 AM EST

but supposedly there is a surge of newly registered voters this year.

Does anyone have any numbers of NRVs(Newly Registered Voters) pre and post primaries and how those numbers compare to the previous presidential elections?

If there is a sharp increase in NRVs this year, I'd be interested in how the statistical trends for previous PEs(presidential elections) match up to this year.

I'm all for registering as many voters as we can and getting as much participation in the process as possible. NRVs historically are not the most reliable voters and I wonder if that will change.

2004 turnout was really pretty massive (none / 0) (#4)
by andgarden on Mon Oct 13, 2008 at 08:48:19 AM EST

and either candidate would have easily beaten the other's party's vote total--in 2000.


That's a painful thought. (none / 0) (#18)
by Fabian on Mon Oct 13, 2008 at 09:19:35 AM EST

I didn't realize we had that many easily misled [polite version] voters in 2004.

So the lousier conditions are, the more people vote?


: Who knows. . . (5.00 / 0) (#19)
by andgarden on Mon Oct 13, 2008 at 09:22:38 AM EST

I'm pretty sure that 3rd world turnout routinely beats ours, thouh.

: Two thoughts (5.00 / 0) (#24)
by Lou Grinzo on Mon Oct 13, 2008 at 09:47:00 AM EST

First, lousy conditions do boost voter turnout. If people are happy they feel they have less reason to express themselves in elections.
Second, I'm not surprised at the number of "misled" first-time voters, given ho0w the Republicans were telling the world that Democrats would take their guns, raise their taxes, and give both the guns and the money to the married gay people next door.

By not picking a strategy and sticking with it... (none / 0) (#5)
by barryluda on Mon Oct 13, 2008 at 08:49:33 AM EST

McCain has painted himself into a corner.

He keeps asking "Who is Obama?" but the real question has become "Who is McCain today?"

Along those lines, I gotta wonder which McCain shows up on Wednesday. I'm confident Obama will be ready no matter what, but I'll be glad when I'm watching all the pundits spin what will most likely be a tie (that goes to the "runner", Obama).

I'm constantly reminded (5.00 / 1) (#7)
by andgarden on Mon Oct 13, 2008 at 08:51:44 AM EST

of this ad from 2000.


: The punditry noise is taken ... (5.00 / 1) (#26)
by Robot Porter on Mon Oct 13, 2008 at 10:04:45 AM EST

with a huge grain of salt by a lot of voters.
Gore still won the popular vote despite aggressive attacks by the media, and ads like the one above.
Hillary was being called erratic shortly before her big primary comeback wins.
Obama's current lead has much less to do with McCain's behavior, and much more to do with the Democratic brand on the economy.
In short, if Obama wins ... blame Clinton.
;)

I have read that LV estimates (none / 0) (#8)
by befuddledvoter on Mon Oct 13, 2008 at 08:54:16 AM EST

undercount the new voters substantially as one of the questions asked and weighed is did you vote in the last presidential election. So, this means the newly enrolled voters tend to be undercounted as "likely voters" for the stats. I think that indicates that the tallies are low for Obama since manny of the newly enrolled will vote Obama.

Gallup is using two LV models (5.00 / 1) (#10)
by Cream City on Mon Oct 13, 2008 at 08:57:55 AM EST

to make you happy. See its site for the LV poll based on the past as predictor and the LV poll with NRV's built into the project.

Then deduct for the ACORN effect -- see previous threads in which we are reassured that all those faked NRVs will not, can not, turn up at the polls.


One year, Republicans (1.00 / 0) (#20)
by MKS on Mon Oct 13, 2008 at 09:25:45 AM EST

were all triumphant that some liberal organization had registered Donald Duck. On and on they went. Voter fraud to the high heavens....Donald Duck, of course, did not vote....

The fraud is on the Democrats who pay for bogus registrations of people who don't vote.....

But you exaggerate the issue--Obama hasn't relied on ACORN at all since the Primaries....


You think ACORN is registering (5.00 / 0) (#23)
by Cream City on Mon Oct 13, 2008 at 09:44:26 AM EST

McCain voters?

Well, it may be registering those who like the Obama/Palin ticket -- one of the most emblematic moments of this campaign yet captured for posterity. Heaven help this country.


well (5.00 / 2) (#27)
by connecticut yankee on Mon Oct 13, 2008 at 10:11:55 AM EST

You didnt address his point. But I understand that addressing it might conflict with your role as prophet of doom.


Of course, I did. (5.00 / 0) (#28)
by Cream City on Mon Oct 13, 2008 at 10:15:31 AM EST

You just don't get the joke, I guess.

Or maybe it's that you disagree with his point that "the fraud is on the Democrats"? I do, to a point -- as the fraud's real costs will come for one and all.


eh (1.00 / 0) (#36)
by connecticut yankee on Mon Oct 13, 2008 at 10:41:40 AM EST

You just took his quote out of context to pursue your shrill nonsense.

The full quote, "The fraud is on the Democrats who pay for bogus registrations of people who don't vote.....".

What is wrong with you?


What is wrong with your comprehension? (5.00 / 0) (#38)
by Cream City on Mon Oct 13, 2008 at 10:47:16 AM EST

The full quote means something different and better to you? Yikes.


well (1.00 / 0) (#55)
by connecticut yankee on Mon Oct 13, 2008 at 11:41:28 AM EST

absolutely it's your problem.

The context of the post is that registration for financial gain doesnt mean anyone is actually going to cast a vote. It's a problem but it doesnt indicate election day fraud. It's likely just greed.

That you cherry picked the post to smear democrats, as usual, says more about you than ACORN does about Obama.


: Which is what I said (5.00 / 0) (#57)
by Cream City on Mon Oct 13, 2008 at 11:45:46 AM EST

all those faked NRVs will not, can not, turn up at the polls
.
There is more to it, of course, in terms of other impacts. But those are detailed in other comments.
So go find another playground, please.

Funny clip (none / 0) (#30)
by Faust on Mon Oct 13, 2008 at 10:28:18 AM EST

but emblematic? I think not. Creating a clip that reveals the stupidity of the electorate is like shooting fish in a barrel.

You just defined (none / 0) (#31)
by Cream City on Mon Oct 13, 2008 at 10:33:36 AM EST

emblematic, of course.


No (none / 0) (#42)
by Faust on Mon Oct 13, 2008 at 11:02:41 AM EST

because I'm arguing that rather than an exemplary illustration of type, it's merely a commonplace.

In my opinion in order for this to be "one of the most emblematic" moments of this campaign, it should be something that reveals a special stupidity that exemplifies all the other stupidities. But moments like the ones generated by the interviewer are standard late-night comedy fodder and are easily generated. (Witness for example the ease with which a supporter of the McCain/Biden ticket was found).

There is nothing special or noteworthy about this clip. It exemplifies nothing. It is not a special noteworthy example of stupidity. It is standard fare.

Or to put it another way: It would be as though I took a quarter out of my pocket and proclaimed: "this quarter is the most emblematic quarter of all other quarters!"


Ah, I see it (none / 0) (#45)
by Cream City on Mon Oct 13, 2008 at 11:06:08 AM EST

now, that the modifier is the prob. Okay, there's an argument that there are no stupidities that are special in this campaign year. And think how depressing it would be to have to review them all to rank them.


: Yes that would be very depressing (none / 0) (#53)
by Faust on Mon Oct 13, 2008 at 11:37:22 AM EST

but I'm sure there would be some good laughs along the way :)

There is still room to worry (none / 0) (#9)
by vector on Mon Oct 13, 2008 at 08:57:28 AM EST

First, I need to apologize. I accidentally posted this earlier, in the Bill Kristol thread, when I meant to post it here.

Anyway, this is all what I want to say:

The Ohio Newspapers' recent statewide Poll has Ohio basically tied - with McCain having slight lead.

Bottom line: I fear that it could be a very long night on November 4, after all.

See article (from yestrday's Cleveland Plain Dealer): http://tinyurl.com/4bpfpl

Poll Internals (PDF format): http://tinyurl.com/4fnkbs

Plus Indiana will be past midnight (5.00 / 0) (#11)
by Cream City on Mon Oct 13, 2008 at 08:59:05 AM EST

again, as the Gary mayor repeats his performance of the primary. It was worthy of Broadway, so why not an encore?


You realize that whatever (5.00 / 1) (#12)
by andgarden on Mon Oct 13, 2008 at 09:00:33 AM EST

direction Indiana goes in, if it's not called until midnight, that's GOOD NEWS for Obama.


especially if it comes down to gary :) (5.00 / 0) (#15)
by sancho on Mon Oct 13, 2008 at 09:10:48 AM EST

Well, that didn't work in the primary (5.00 / 0) (#16)
by Cream City on Mon Oct 13, 2008 at 09:13:38 AM EST

but the Gary mayor has had time to perfect it since. I hope he hasn't had time to get a better script with more plausible excuses, though, or there goes one of the best political laff riots since Plunkitt of Tammany Hall.

Subtitled A Series of Very Plain Talks on Very Practical Politics, Delivered by Ex-senator George Washington Plunkitt, the Tammany Philosopher, from His Rostrum -- the New York County Court House Bootblack Stand, it is a worthy read today. See the Table of Contents:

Chapter 1. Honest Graft and Dishonest Graft
Chapter 2. How To Become a Statesman
Chapter 3. The Curse of Civil Service Reform
Chapter 4. Reformers Only Mornin' Glories
Chapter 5. New York City Is Pie for the Hayseeds
Chapter 6. To Hold Your District: Study Human Nature and Act Accordin'
Chapter 7. On The Shame of the Cities
Chapter 8. Ingratitude in Politics
Chapter 9. Reciprocity in Patronage
Chapter 10. Brooklynites Natural-Born Hayseeds
Chapter 11. Tammany Leaders Not Bookworms
Chapter 12. Dangers of the Dress Suit in Politics
Chapter 13. On Municipal Ownership
Chapter 14. Tammany the Only Lastin' Democracy
Chapter 15. Concerning Gas in Politics
Chapter 16. Plunkitt's Fondest Dream
Chapter 17. Tammany's Patriotism
Chapter 18. On the Use of Money in Politics
Chapter 19. The Successful Politician Does Not Drink
Chapter 20. Bosses Preserve the Nation
Chapter 21. Concerning Excise
Chapter 22. A Parting Word on the Future Party in America
Chapter 23. Strenuous Life of the Tammany District Leader

that plunkitt (5.00 / 1) (#29)
by sancho on Mon Oct 13, 2008 at 10:15:54 AM EST

primer looks great, cream. i'll have to find that and work it into my classes!


: I read it an undergrad course (none / 0) (#33)
by Cream City on Mon Oct 13, 2008 at 10:35:51 AM EST

many (ahem) decades ago and still have it -- as it's still one of the most memorable texts for me, one of the first that gave me the long view (always a comforting view, in a way) of our political history.
Or we could just teach our students the lesson in the lyrics by that great historian, Cole Porter: Everything old is new again.

I love this ... (5.00 / 2) (#37)
by Robot Porter on Mon Oct 13, 2008 at 10:46:13 AM EST

chapter:

Chapter 12. Dangers of the Dress Suit in Politics

Here's the intro:

PUTIN' on style don't pay in politics. The people won't stand for it.
If you've got an achin' for style, sit down on it till you have made
your pile and landed a Supreme Court Justiceship with a
fourteen-year term at $l7,OOO a year, or some job of that kind.
Then you've got about all you can get out of politics, and you can
afford to wear a dress suit all day and sleep in it all night if you
have a mind to. But, before you have caught onto your life meal
ticket, be simple. Live like your neighbors even if you have the
means to live better. Make 'the poorest man in your district feel
that he is your equal, or even a bit superior to you.

Very amusing. But also a political rule few politicians follow anymore. There used to be many who did.


Politicians and (5.00 / 0) (#39)
by Cream City on Mon Oct 13, 2008 at 10:51:39 AM EST

ballplayers. Imagine, I grew up in a middle-class neighborhood with professional baseball players nearby -- Warren Spahn was the Little League coach for one of my brothers. Today, they come in from swank burbs in their limos and hit up little kids for twenty bucks for an autograph.

Of course, I never imagined a campaign that would charge for buttons and list the buyers as donors -- an amazing gimmick that worked, and with the media reporting it, too. I bet Plunkitt would have loved it. :-)


: Plunkitt's key phrase: (none / 0) (#44)
by Robot Porter on Mon Oct 13, 2008 at 11:04:55 AM EST

Live like your neighbors even if you have the means to live better.

Um, ok (none / 0) (#17)
by andgarden on Mon Oct 13, 2008 at 09:16:08 AM EST

You missed my point, apparently.


: Poblano says that (5.00 / 0) (#22)
by MKS on Mon Oct 13, 2008 at 09:32:10 AM EST

the Obama ground game can be measured by whether they can't call Indiana when its polls close, or shortly thereafter. Makes sense to me....

So, you have picked upt the voter fraud (none / 0) (#21)
by MKS on Mon Oct 13, 2008 at 09:28:52 AM EST

theme this morning....Isn't that what Republicans are pushing today....It really shows how demoralized Obama's opponents really are....When you start complaining about the refs, or that the other side is cheating, you ain't winning.


I don't want voter ID (5.00 / 1) (#25)
by Cream City on Mon Oct 13, 2008 at 09:50:37 AM EST

but am demoralized to find that, despite fighting it off legally in my state over and over, we now have it in practicality because of ACORN -- as even we regular, longtime voters in this state are now advised to bring ID to deal with expected problems at the polls.

Yeh, I find that demoralizing. You like voter ID, then be happy for us from afar.

Up close, it sucks. So do the city offices that will be closed or short-staffed again as workers are pulled over to the election offices to try to get the rolls ready in time. We went through that last time, and this is getting tiresome and costly to do time after time. And the GOP now has more ammo to take away a lot of laws here, like same-day registration, too.


You are blaming ACORN (none / 0) (#46)
by samtaylor2 on Mon Oct 13, 2008 at 11:06:52 AM EST

For voter ID laws?

Generally speaking ACORN does amazing work. Registering thousands if not upwards of a million low income Americans is good work. Don't demonize a group for someone trying to up their registration numbers for the day, before the group does their work of sorting out the nonsense.

The mistake that ACORN makes is that they need to "hire" these people as independent contractors, and then put more money and effort into taking out the nonsense.


: See the coverage (5.00 / 0) (#50)
by Cream City on Mon Oct 13, 2008 at 11:18:00 AM EST

and editorials at jsonline.com. Yes, that is how its transgressions are being used in my city and state.
Of course, it's part of a larger context of years of overblown coverage of such stuff from other groups and idiots, too. But it's hard to blame the media for covering stories like not just a Dem but a Dem candidate who voted twice, or a Dem Congresswoman's son who got caught vandalizing GOP vans, or . . . well, ACORN is just the largest group of a lot of lovelies who over-reached -- and in a solidly blue city, for pity's sake -- so we are getting the over-reaction.

a new low for you, Cream (none / 0) (#67)
by Iris on Mon Oct 13, 2008 at 03:06:19 PM EST

ACORN is responsible for Voter ID laws? That is the funniest thing I have ever heard.

People who push for voter ID laws are responsible for them. And those just happen to be Republicans, the people whose agenda you are servicing when you spread falsehoods and misleading notions about an important group like ACORN.

You really, REALLY need to quit reading PUMA websites.


: Iris, you really need to read (5.00 / 0) (#68)
by Cream City on Mon Oct 13, 2008 at 03:24:14 PM EST

coverage and editorials of this at jsonline.com. That is the real world in which I live, not the la-la land of yours where actions don't have consequences.
And bless your heart, but I don't read PUMA sites. Your comments provide all the jollies I can handle.

McCain has a very good shot at Ohio imo. (5.00 / 1) (#32)
by Faust on Mon Oct 13, 2008 at 10:33:45 AM EST

But He needs much more than just Ohio at this point.


Eary Night (none / 0) (#14)
by robrecht on Mon Oct 13, 2008 at 09:09:54 AM EST

IF we're doing well in Virginia or especially in Florida.


If we win VA ... (none / 0) (#34)
by Robot Porter on Mon Oct 13, 2008 at 10:38:45 AM EST

you can put on your jammies, if we win FL you can go to bed. No matter what the other states look like, a FL win means Obama has won.

There's is no way a Republican can win the Presidency right now without FL.


: Florida (none / 0) (#48)
by CoralGables on Mon Oct 13, 2008 at 11:08:18 AM EST

Just keep an eye on the individual Counties as they roll in. Part of the state is Central Time (and not the part that is loaded with Dems)
Broward is pretty efficient for a large county, and will get their results up while some small counties lag. If Obama has the lead as Broward rolls in, the State may turn blue. If Obama isn't in the lead after Broward results are tallied, start looking elsewhere.

: i remember going out to celebrate (none / 0) (#60)
by sancho on Mon Oct 13, 2008 at 12:43:27 PM EST

after florida was called for gore. and then i saw in a bar the bush family announcing that the florida call was wrong and would be updated shortly.
and then i thought i saw all of us falling through the rabbit hole.
i'm hopeful but still wary until the dem nominee is sworn in.
but, yeah, florida ought to do it. or ohio.

to: larry in nyc (none / 0) (#35)
by oculus on Mon Oct 13, 2008 at 10:40:31 AM EST

did you base you prediction on LV, RV, or what?

Some Monday State Polling (none / 0) (#41)
by CoralGables on Mon Oct 13, 2008 at 11:00:51 AM EST

Pennsylvania Obama +13 (Morning Call)
Pennsylvania Obama +12 (Marist)
Ohio Obama +4 (Marist)
North Dakota Obama +2 (Forum Poll/MSUM)

I wish pollsters wouldn't sit on their results (none / 0) (#47)
by andgarden on Mon Oct 13, 2008 at 11:07:11 AM EST

over the weekend.


: Some Pollsters... (2.00 / 0) (#49)
by CoralGables on Mon Oct 13, 2008 at 11:09:36 AM EST

like some republican presidential candidates, don't work weekends

One More (none / 0) (#61)
by CoralGables on Mon Oct 13, 2008 at 01:20:42 PM EST

Missouri Obama +8 (Survey USA)


: That is more interesting (none / 0) (#62)
by Cream City on Mon Oct 13, 2008 at 01:37:53 PM EST

and indicative than the Eastern states -- Penn, Fla, Ohio -- I think. They may clinch it early, sure . . . but if not, then watch the heartland, and especially the crucial Mississippi River "coast" states that add up to many electoral college votes.
So now, watch Minn and see if there is settling down of the polls a few days ago that had that state from 1 point up to 18 points up for Obama on the same day or something!
I really hope we don't have to wait all the way to the Pacific Coast . . . and then watch one of the Eastern states flipped again to the GOP in the wee hours of the morning. A decisive win by the time we get to the midsection of the country would be nice for a change.

Late Day Adds (none / 0) (#71)
by CoralGables on Mon Oct 13, 2008 at 05:32:12 PM EST

Georgia McCain +8(Survey USA)
New York Obama +33(Survey USA)
New Jersey Obama +15(Survey USA)
Florida Obama +5 FoxNews/Rasmussen
Missouri Obama +3 FoxNews/Rasmussen
North Carolina (TIE) FoxNews/Rasmussen
Ohio Obama +2 FoxNews/Rasmussen
Virginia Obama +3 FoxNews/Rasmussen


Gallup ticks up on all measures (none / 0) (#58)
by andgarden on Mon Oct 13, 2008 at 12:12:31 PM EST

especially their new LV screen.

Which of their LV models do you like? (none / 0) (#63)
by Cream City on Mon Oct 13, 2008 at 01:38:38 PM EST

Or do you go with RCP's decision to just average both?


I like the least manipulated data (none / 0) (#64)
by andgarden on Mon Oct 13, 2008 at 01:59:15 PM EST

available: registered voters.


Interesting. Just so I understand (5.00 / 0) (#65)
by Cream City on Mon Oct 13, 2008 at 02:11:27 PM EST

that (I've asked this before but haven't seen a clear answer, as it seems to vary in some polls) -- that means previous voters and potential voters who have registered for this election to vote for the first time? So those would be people who actually never went to the polls but say they will this time, but those are not counted as likely voters?

I guess this really gets a bit boggled for me in a same-day registration state. Plus, some polls seem to use people registered by party, and I don't see what that means in states that don't register by party.

I guess the problem is that definitions vary?


Definitions do vary, but I think (none / 0) (#66)
by andgarden on Mon Oct 13, 2008 at 02:14:56 PM EST

for at least 90% of the country, the interesting questions are 1) Are you registered to vote?; and 2) Do you intend to vote on November 4th?

Dealing with same day registration is tricky, but I think the pollsters should know which states have it, and modify question 1 as appropriate.


: That's what I'd think, but (5.00 / 0) (#69)
by Cream City on Mon Oct 13, 2008 at 03:28:21 PM EST

none of them seem to address it -- I've looked. And when I've been surveyed, many times since I'm in a swing state, and one with same-day registration, none of the questions ask about it.
Only one survey even acknowledged other laws in my state, laws by which all pollsters calling us are bound. So I've been able to make a lot of use of my bookmark to the state agency that enforces it.
Other questions also have been, at times, unanswerable. All in all, I am not impressed with the national pollsters calling us, as most don't know the state law -- or the state. That could be why they got my state wrong last time, too, when most of them called it for Bush. (Whew.)