Exit Poll: Turnout and Ballot Spoilage Could Have Swung Popular Vote

[Bill Bored]

This is a continuation of this thread:
<http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=225847&mesg_id=225847&page=>

Comparing the latest exit poll report (1AM) from the Wash. Post
<http://www.washingtonpost.com/wp-srv/politics/elections/2004/graphics/exitPolls.html>
to the final one released by Mitofsky at 1:30 PM, a major adjustment was made in party weightings by changing Dem/Rep 38/35 to Dem/Rep 37/37. Obviously this implies an equal voter turnout between the two major parties, which was not assumed in any of the earlier versions of the poll.

From what I can see, all other weightings, including gender, age, race, national region and religion, were changed by 1% or less from the last unadjusted version of the poll to the final version and had a negligible effect on the candidates' margins.

The party weightings alone however, which simply reflect Democratic and Republican voter registration and turnout, would have reduced Kerry's popular vote lead from the 3% he enjoyed throughout the day, to a scant 0.6% in the final tally. While this alone didn't give Bush a popular vote victory, it would have put him closer to the margin of error of the poll so that such factors as ballot spoilage, uncounted provisional ballots and machine "glitches" could have eroded Kerry's remaining 0.6% margin, and given Bush a plurality within the poll's 1% margin of error.

For this hypothesis to be correct, we'd need to know one thing:
Was the Republican and Democratic turnout, in actual millions of votes, really about equal? This is where voter suppression comes in, because there are more of us than there are of them.

Nationally, we know about Registration Fraud against Democrats, which in effect "aborts" new voters thus reducing turnout. In Ohio, we know about voter suppression by various means, including fewer machines in Democratic precincts, changing precincts at the last minute, uninformed election workers accepting provisional ballots in incorrect precincts only to have them rejected at counting time, etc. These tactics of disinformation and chaos in Ohio can all be attributed to Mr. Blackwell. While most of the provisional ballots in Ohio were counted (23% were not), nationwide I believe this figure is somewhat less. And there are Cliff Arnebeck's and the Glibs' other Ohio allegations. Florida, as usual, is a lost cause but for the record, we've heard about absentee ballots getting lost, spindled, mutilated and destroyed, esp. in Teresa LePore's county. Who knows what else went on down there among the Dixiecrats, not to mention the ReShrublicans, in Jebville. IMHO, FL just can't be taken seriously when it comes to elections. Maybe Howard Dean can sort it out, along with the other 49 states he thinks we can win in, but he's a better man than I am!

If you believe that the above factors could have equalized voter turnout by depressing Democratic turnout, you don't need any other evidence of large scale hacking of the vote to explain the exit poll discrepancy in my opinion. Suppression, spoiled ballots and the uncounted provisionals probably account for Bush's lead, although of course, the vote is still unverifiable in many states due to the lack of a paper trail and this must be changed as soon as possible.

It would be helpful now to find an independent source of voter turnout , besides this one exit poll, to confirm this theory. Is there one, or is the whole country dependent on Warren Mitofsky's little "adjustments?"

[davidgmills]

1. Suppression should not be part of these stats

If you mean suppression as in people who did not get to vote at all. They would not have been exit polled.

[Bill Bored]

2. I thought the same thing, but the polls show an equal number of Rs and Ds.

We know in fact there are more Ds. So in this way, suppressed Democrats are reflected (by not being counted) in the poll.

[Mistwell]

3. Why?

Why do we know in fact that there were more Democrats? We suspect it, but the whole point of the thread seems to be...do we actually know it?

[Bill Bored]

6. I think we know there are more registered Dems. But

the poll says they turned out equally to Repubs to vote. Hence suppression.

The other eplanation is that Rove got out the vote better than we did. I still don't buy that. Dems worked really hard this year! Unless the exit poll releases with the Kerry lead motivated Shrub's base, which I've always said.

[davidgmills]

4. The democrats were counted as being ahead

Until the "corrections" were made. So they weren't surpressed in that sense. Am I missing something?

[Bill Bored]

5. This is why I was asking if there are INDEPENDENT numbers for turnout.

Why should they adjust the proportion? They could have just shown more votes for Bush, but the equal turnout explains why. The only other source of vote count is the actual vote count, but there may be another source of turnout data (without showing how they actually voted).

[davidgmills]

7. I know that before the election

There was much speculation and argument about what the ratio of dems to repubs should be when polling. Zogby had a formula of about 39 dems to 35 repubs. Unadjusted polls say those numbers were right. Other pollsters thought that repubs had closed the gap and had the two even at 36-38 a piece. I remember arguing with some freepers about Gallup being crazy because he kept polling more repubs than dems in his polls and was getting these huge leads for Bush. They kept insisting there was no gap and some argued that Republicans had pulled ahead.

For future polling this is ominous. It will allow all pollsters in the future to poll based on a 50/50 split when the actual polls showed otherwise.

I just think Mytofsky caved into freeper arguments and used that as a basis to match polls with actuals.

[tritsofme]

8. Zogby pegged party ID at 2000 levels

Pollsters like Gallup and Pew used a random sample and did not weight their polls based on party ID of that random sample. You can't oversample a group if the sample is random.

Many people, including myself, before the election saw a shift toward the GOP in terms of party ID before the election, and this shift did not come as much of a surprise.

There's no problem to future polls, regardless of problems in the 2004 exit polls, if they use a random sample instead of pegging their data to the political situation of four years ago.

[Bill Bored]

9. So are you saying Mitofsky under-sampled Repubs before he adjusted? n/t

[tritsofme]

10. We don't know the reasons for the adjustments

but we do know that Mitofsky warned his clients not to rely on the preliminary data.

[davidgmills]

11. My conclusion was that the shift was spin

My suspicion now is that any shift was due to past election fraud especially 2002. Freepers cited 2002 as evidence of the shift.

I saw the inital raw exit polls and my thought was: "so much for shift."

By the way I am almost sure Rassmussen was doing an even split in his polling.

To get a surplus of democrats in the exit polls, Mytofsky would have had to go to some heavily democratic areas and intentionly skew the results. I have never heard as an excuse for the exit polls that Mytofsky intentionally over plolled democratic precints. The only excuse given for the gap between dem and repub numbers is that the repubs refused to talk to pollsters. Freeman has pretty much tore taht excuse a new one. The other excuse was gender which has even been disavowed by Mytofsky, I beleieve. But there was never any mention of overpolling dem precincts.

[zalinda]

12. what about new registrations

If you assume that 2000 was correct, (yeah, right) then that would be about 50-50, saying that the Nader factor was a moot point. Didn't the dems out register the repugs? I read/heard some where that it was like a 3 to 1 factor. Where did they all go?

Kalinda

[Darkhawk32]

13. They went into the self-eating code. n/t

[spooked911]

14. I have been wondering the exact same thing!

[m.standridge]

15. but factor in southern Democrats

and Democrats associated with Evangelism who turned out heavily to vote against gay marriage in several states.
Media has reported that African-American turnout for Bush was higher than in 2000, too.
And don't forget those pre-election polls of the last couple of days.
One thing intriguing about them, to me, is how they usually showed a back and forth race at the national level, with Bush more often having a lead (barely) than not, but with Kerry carrying Ohio and the other states necessary for him to win in the Electoral College, more often than not, too.
In many ways, it began to look to me, even before the election, that Kerry might be the first Democrat to win in the Electoral College only.
There's still that possibility.
Or, that it was very close--too close to call as to the Popular vote, but that Kerry is doing considerably better, in reality, in the Electoral College than it is looking like right now.

[Bill Bored]

16. Dem/Rep Turnout Weighting Changes by Region

Here are the numbers comparing the Dem/Rep ratio at 7:33 PM 11/2 to 1:24 PM 11/3 by region:


11/2 11/3
East
D 42 41
R 28 30

MidWest
D 38 37
R 38 38

South
D 37 35
R 41 42

West
D 37 34
R 34 36

National
D 38 37
R 36 37


We can clearly see that Mitofsky adjusted these in all 4 regions, reducing his estimates of Dem turnout regardless of how people actually voted. Once again, this accounts for much of the shift to Bush, although he also had to use some actual vote swapping "adjustments" to fully conform to the official vote counts which would have included ballot spoilage, etc.

The shift in turnout was greatest in the west (Dem = -3%/Rep = +2%), which was actually enough to REVERSE it in this region. This is consistent with the notion that the early release of these polls may have helped stimulate Repub turnout. No other region had their turnout numbers reversed, although timezone may have also had something to do with this. Looking at the earliest release at 4PM might shed some light on this, but I don't have time right now.

The big question remains, are these turnout numbers believable?

[L. Coyote]

17. Uncounted votes is a real concern. This problem must be fixed. n/t