A Conversation about the 2008 Election (TIA)

[tiptoe]

A Conversation about the 2008 Election

TruthIsAll      source: http://www.geocities.com/electionmodel/Conversation2008.htm

Dec. 2, 2008


What’s the 2008 recorded vote count?

Obama leads by exactly 9.0 million votes: 68.33 – 59.33m  (52.76 – 45.81%)  with 129.520m recorded votes counted to date.

Who voted?

Two groups: Those who ...

1)  Voted in 2004

 

(Returning Kerry, Bush and other, third-party 2004 voters)

2)  Did Not Vote in 2004

 

(Newly-registered-for-2008 voters and others who voted prior to 2004)

Can we estimate the number of 'New' and Returning-Election-2004 voters in 2008?

There were 122.3 million official recorded votes in 2004. Bush won by 62 – 59m  ( 50.73 – 48.27% ).
Assuming a 4-year voter mortality rate of 5 percent, 6 million died.
Therefore, about 116m who voted in 2004 were alive to vote in 2008.

• About 95% -- or 110 million -- turned-out to vote in 2008. Assuming the official 2004 recorded vote shares,
  50.73% (56m) are returning-Bush-voters and 48.27% (53m) are returning-Kerry-voters.
  
  
• ‘New voters’ in 2008, then, would presently total 129.52m – 110m, or approximately 19.5 million.

You’re assuming the 2004 Recorded vote was equal to the True Vote. Researchers have concluded that Kerry won by 8-10m. What about that?

Let’s not get bogged down by a discussion of election fraud. Let’s just accept that the 2004 Recorded vote was the True Vote. Ok?


OK. Do we know how the two groups voted in 2008?

According to the Final 2008 National Exit Poll (find “2004”):

 

Obama won

and

McCain won

Of the

53m

 

(40.9%)

 

returning-Kerry-voters

 

89%

=

47.2m

 

9%

=

4.8m

Of the

56m

 

(43.2%)

 

returning-Bush-voters

 

17%

=

9.5m

 

82%

=

46.0m

Of the

1m

 

( 0.8%)

 

returning-Other-voters

 

66%

=

0.6m

 

24%

=

0.2m

Of the

19.5m

 

(15.1%)

 

'new' (DNV in 2004)

 

71%

=

13.8m

 

27%

=

5.3m

129.5m

 

(100.0%)

 

OK, then, using the 2004 official Recorded vote count and shares, the 2008 official Recorded vote total and the 2008 Final National Exit Poll vote shares, you are saying that Obama must have won by:

71.1m – 56.3m     ( 54.9 – 43.5% )

But that does not include Uncounted votes, which are 70-80% Democratic.
Let’s assume the following Final 2008 NEP Scenarios for returning voters and a 3% uncounted vote rate:

• Bush 50.7–48.3% official 2004 recorded vote margin:
  Obama’s True vote share was 55.5%   (a 16.9m vote margin).
  
  
• Kerry’s 52–47% unadjusted state exit poll (Edison-Mitofsky’s ‘WPE') margin:
  Obama’s True vote share was 57.7%   (a 22.8m vote margin).

I thought we agreed that this discussion was to be based on the official Recorded vote.
The Final NEP was matched to the recorded vote: Obama’s vote share was 52.76% (9.0m vote margin).


But we just calculated a different result based on official Recorded Votes and Final NEP. How could that be?

The 2008 Final National Exit Poll used a different returning-2004-voter MIX in order to ‘force’ a match to the recorded vote. It indicated that returning-Kerry-voters comprised 37% of the recorded vote; Bush-voters 46%; Other 4%; Did Not Vote 13%.


But how could 46% of 129.52m (59.6m) have been returning Bush voters?

Only 59m of the 62m Bush-voters were alive in 2008 (approximately 3million died), and about 56 million (95%) voted in 2008.

And just 37% (47.9m) were returning Kerry voters?

How could Bush-voters outnumber Kerry-voters by 11.7million, when Bush’s official 2004 vote margin was only 3.0 million?

And 4% of the 2008 electorate (5.2m) consisted of third- party 2004 voters?

The official recorded vote count shows there were only 1.2m third-party voters in 2004.

The Final NEP is always forced to match the recorded vote count.
There are two possibilities:
a) returning third-party voters misspoke; they did not want to admit that they voted for Bush in 2004 or
b) the Final NEP was forced to match the 2008 recorded vote; the returning-2004-voter mix and/or 2008 vote shares had to be “adjusted”.


What about the unadjusted 2008 state exit polls?

We don’t have those numbers yet. Exit Pollsters Edison-Mitofsky should release their detailed report in a few months.

Didn’t they report that Kerry won the unadjusted exit poll by 52–47%, based on the average within precinct discrepancy (WPE)?

Yes, but there’s a catch.

Are you referring to the exit pollsters claim that the discrepancy was due to Bush voter reluctance (rBr) to be interviewed? Wasn’t that theory refuted by other statistics from the exit pollsters themselves, including the Final NEP 43/37 Bush/Gore returning voter mix?

Yes, it was. But the real reason was “false recall” on the part of Gore voters who were exit polled; they misspoke when they indicated they voted for Bush.

And why would they do that?

They wanted to identify with the “winner” of the 2000 election. “False recall” is still a viable hypothesis.

But Bush had a 48% approval rating in 2004, and Gore was the official winner of the popular vote. “False recall” wasn’t a viable hypothesis then, and it’s not one now. Are you claiming that the 46/37 Bush/Kerry mix in the 2008 Final NEP is due to returning-Kerry-voters who indicated that they voted for Bush in 2004? Bush had a 22% approval rating in 2008. What would be the motivation of Kerry voters who voted for Obama to indicate that they voted for Bush in 2004?

Maybe this time they just really forgot that they voted for Kerry. Not that they wanted to identify with Bush, mind you. Just forgot.

Oh. How do you explain Obama winning the final 8.3 million votes (absentee, provisional) by 59-39%?

What does that prove? He won the first 121m by 52.7-46.0%, and 8.3 million is too small a sample size to draw any conclusions

But what if the recorded vote is once again fraudulent? The evidence indicates Obama won by 16-22m votes, not the 9m recorded.

There you go again. Back to your old conspiracy theories, just like 2000, 2002, 2004, and 2006. Obama won. Get over it.

 

[Sancho]

1. The good thing about TIA is that he is still raising a reasonable question...

It is true that TIA may have exaggerated probabilities associated with random binomial distributions and also that nonignorable nonresponse (Howard Wainer's name for "reluctant respondent) was overused by pollsters; but the fact that adjustments don't add up to intuitively logical results is a red flag that there is an unknown factor in the reported results. The anecdotal reports of DRE switching, tabulator hacking, caging, missing mailed ballots, manipulated ballots (butterfly, etc.), miscounts, weird undervotes, inconsistent results with local elections, registration difficulty, etc. seem to allow for the possibility of election manipulation on a planned scale.

IF the pollsters wanted to provide evidence at the precinct level or aimed at finding election manipulation; they could do that and they simply don't want to...and it is a waste of time to continue to talk about they way that pollsters sample again and again. Major NEP polls are NOT trying to examine why the results don't make sense. My theory is that some pollsters and some election supervisors and possibly Karl Rove already know why the results don't add up.

Here in Florida it happened again. There is no reasonable explanation that 250,000 new Democratic registered voters (many minority) and a record turn out of those same voters would go in a booth and vote for Obama and also vote for a GOP hack for senate or a GOP sponsored amendment?! Senile and stupid is one thing, but these discrepancies are practically impossible to accept in many cases. I'd bet that any reasonable exit polls at the precinct level, even with small samples, would suggest that SOME elections were very, very questionable results. There is no reason for exit polls to fail to follow up on that type of evidence after all the controversies.

[truedelphi]

2. You Say:

Here in Florida it happened again. There is no reasonable explanation that 250,000 new Democratic registered voters (many minority) and a record turn out of those same voters would go in a booth and vote for Obama and also vote for a GOP hack for senate or a GOP sponsored amendment?! Senile and stupid is one thing, but these discrepancies are practically impossible to accept in many cases. I'd bet that any reasonable exit polls at the precinct level, even with small samples, would suggest that SOME elections were very, very questionable results. There is no reason for exit polls to fail to follow up on that type of evidence after all the controversies.


####

Florida is going to be a real tough nut to crack in terms of all the Dems who are really Repugs handling the situation at polling places.

Can I ask this - when the votes were somewhat recounted in Florida in 2000, did the vote counters look at the ballots as a whole, or just at the Presidential entries??