1988-2000 Exit Poll Discrepancies were due to Uncounted Votes; but 2004 Differed in Kind and Scope

[tiptoe]

Uncounted and Switched Votes

TruthIsAll      source: Further Confirmation of a Kerry Landslide, Ch. 19

1988-2004: Uncounted Votes and Exit Poll Discrepancies

Uncounted votes have steadily declined as percent of total votes cast — from 10.4% in 1988 to 2.7% in 2004. When uncounted votes are included in order to derive total votes cast for the five elections from 1988–2004, the adjusted Democratic vote is within 0.1% of the unadjusted exit poll share.

Comparing the adjusted vote to the aggregate state exit polls and the recorded vote:
(2-party exit poll share in parenthesis)

 

Year

 

Democrat

 

RECORDED
VOTE-COUNT

UNCOUNTED
ADJUSTED

 

UNADJUSTED
EXIT POLL

AVERAGE

 

 

 

46.9 %

48.9 %

 

48.8 %

 

(52.7)

 

2004
2000
1996
1992
1988

 

Kerry
Gore
Clinton
Clinton
Dukakis

 

48.3
48.4
49.2
43.0
45.6

49.0
49.7
51.4
45.7
48.7

 

51.8
49.4
50.2
45.7
46.8

 

(52.3)
(51.4)
(55.8)
(56.8)
(47.3)

Look at this ?click">graph:

In every one of the last five elections the unadjusted Democratic EXIT POLL share exceeded the RECORDED vote. But which of the five elections stands out from the rest? Only in 2004 did the ADJUSTED vote come up short of the EXIT POLL share.

The 2004 exit poll discrepancies were different in kind and scope from those of the prior four elections.
Unlike 1988-2000, the 2004 discrepancies cannot be explained by uncounted votes alone.

The analysis shown below indicates that given Census 2000/2004 estimates of total votes cast and 12:22am National Exit Poll “Voted 2000” shares, Bush needed 21.5% of returning Gore voters to match his recorded vote! Even using Final NEP vote shares, Bush needed 18.1% of returning Gore voters! The Final was 'forced' to match the recorded vote using mathematically impossible weights and by sharply increasing Bush vote shares from the 12:22am NEP.

There are some exit poll critics who claim that the large (5.4 WPE) 1992 exit poll discrepancy proves that the 2004 discrepancy (7.1 WPE) was not unique, and, therefore, to conclude that the election was stolen based on exit poll results is "crap" and "bad science". After all, they say, there were no allegations of fraud in 1992. They fail to mention or are unaware of the fact that in 1992 Clinton beat Bush I by 44.9-39.1m (43.0-37.4%), but that 9.4m votes were uncounted, of which 70–80% were Democratic ( 75% Dem / 25% Rep are used in the model). When uncounted votes are included, the 52.0–41.5m adjusted vote (45.7–36.4%), exactly matched Clinton’s unadjusted exit poll.

From 1988-2000, after the uncounted adjustment, there was a 0.85% average Democratic exit poll discrepancy and 2.9 WPE. In 2004, after the 3.4m uncounted vote adjustment, the exit poll discrepancy was 2.8% and Bush's margin was reduced from 3.0m (62.0–59.0) to 1.3m (62.9–61.6).  But ...

Uncounted votes were only one component of Election Fraud 2004.  The Election Calculator Model determined that approximately 5 million votes were switched from Kerry to Bush.

Recorded and Uncounted Votes

Year

 

States Reported
VOTES COUNTED

VOTES COUNTED

Dem

Rep

Other

 

COUNTED-VOTE  Shares

Dem

Rep

Other

 

WPE

 

VOTES
UNCOUNTED

~75%
Dem

~25%
Rep/Oth

AVERAGE

 

104.0

48.8

47.9

7.2

 

46.9%

46.0%

7.1%

 

3.76

 

7.5

5.7

1.9

 

2004
2000
1996
1992
1988

 

122.3
105.4
96.3
104.4
91.6

59.0
51.0
47.4
44.9
41.8

62.0
50.5
39.1
39.0
48.9

1.2
4.0
9.7
20.4
0.9

 

48.3%
48.4%
49.2%
43.0%
45.6%

50.7%
47.9%
40.7%
37.4%
53.4%

1.0%
3.8%
10.1%
19.6%
1.0%

 

7.09
2.01
1.93
5.40
2.38

 

3.4
5.4
8.7
9.4
10.6

2.6
4.0
6.5
7.1
7.9

0.9
1.3
2.2
2.4
2.6

 

Total Votes Cast vs. Exit Poll

Year
Adjusted

 

Census Reported
VOTES  CAST

COUNTED+UNCOUNTED

Dem

Rep

Other

 

UNCOUNTED-Adjusted Sh

Dem

Rep

Other

 

EXIT POLL
Dem

 

post-Adj
Diff

 

EP2pty
Dem

 

UNCTD
/ CAST

AVERAGE

 

111.5

54.5

49.8

7.2

 

48.9%

44.6%

6.5%

 

48.8%

 

0.1%

 

52.7%

 

7.0%

 

2004
2000
1996
1992
1988

 

125.7
110.8
105.0
113.9
102.2

61.6
55.0
54.0
52.0
49.8

62.9
51.8
41.4
41.5
51.5

1.2
4.0
9.7
20.4
0.9

 

49.0%
49.7%
51.4%
45.7%
48.7%

50.0%
46.8%
39.4%
36.4%
50.4%

1.0%
3.6%
9.2%
17.9%
0.9%

 

51.8%
49.4%
50.2%
45.7%
46.8%

 

-2.8%
0.3%
1.2%
0.0%
1.9%

 

52.3%
51.4%
55.8%
56.8%
47.3%

 

2.74%
4.86%
8.31%
8.48%
10.37%

 

The Census Bureau website states the following:

“The data are from the November 2004 Voting and Registration Supplement to the Current Population Survey (CPS). Statistics from surveys are subject to sampling and non-sampling error. The CPS estimate of overall turnout (125.7 million) differs from the “official” turnout, as reported by the Clerk of the House (122.3 million). For further information on the source of the data and accuracy of the estimates, including standard errors and confidence intervals, go here.

The difference between the 2004 recorded vote total and the U.S. Census estimate is 3.45m votes. The Census survey margin of error is 0.30%. According to investigative reporter Greg Palast, government records show that 3.006m votes were uncounted, comprised of 1.389m spoiled, 1.091m provisional and .0.526m absentee ballots.

Note that the 3.4m estimated difference is a net figure. In 13 states the official vote exceeded the Census estimate by 730,000 votes. The largest discrepancies were in Florida (238k), Ohio (143k) and Tennessee (118k). Apparently more votes were padded than suppressed in the 13 states. But the net discrepancy does not contain the necessary information to calculate the actual number of uncounted and padded votes.

There are many combinations which would provide the net number. For example, of the 3.4m vote difference, 4.0m could have been suppressed (uncounted) and 0.6m padded (stuffed). Or 5.0m suppressed and 1.6m padded. In Florida, the least onerous case would be if the total 238k discrepancy was due to vote padding. But it is far more likely that vote padding occurred in Republican districts while vote suppression took place in heavily Democratic minority districts. The net 238k difference could have been due to 338k padded and 100k suppressed votes.

The Election Calculator

2000:  US Census Votes-Cast and State Records of Votes-Counted
 

 
 

2004 Calculated True Vote

 True  'Voted in 2000'  Mix 

 
 

12:22am NEP  ('13047')
'Voted in 2000' Shares

Total Votes
Cast in '00

Gore
Bush
Nader/Other

110.8
million

 

Recorded
Vote-Count

51.00
50.46
3.96

105.42

 

Uncounted
Allocation

4.04
1.08
0.27

5.38

 

Vote-Count
Adjusted

55.04
51.53
4.23

110.80

 

Deaths

2.72
2.48
0.21

5.41

 

Alive

52.32
49.06
4.02

105.39

 

Est '04 Turnout
of 'Voted 2000'
DNV
95%
95%
95%

100.13

 

Total Votes
Cast in '04

25.62
49.70
46.60
  3.82

125.74
True Vote

 

'Voted 2000'
Weight

20.4%
39.5%
37.1%
3.0%

100.0%
125.74

 

Kerry

57%
91%
10%
64%

53.2%
66.9

 

Bush

41%
8%
90%
17%

45.4%
57.1

 

Other

2%
1%
0%
19%

1.37%
1.73 

 

Sensitivity  Analysis
Kerry National Vote

Kerry Share of
Gore Voters

Share of New Voters (DNV in 2000)

53.2%

53.0%

55.0%

57.0%

59.0%

61.0%

95%
93%
91%
89%
87%

54.0%
53.2%
52.4%
51.6%
50.8%

54.4%
53.6%
52.8%
52.0%
51.2%

54.8%
54.0%
53.2%
52.4%
51.7%

55.2%
54.4%
53.6%
52.8%
52.1%

55.6%
54.8%
54.0%
53.3%
52.5%

 

Kerry Margin (millions)

9.87

53.0%

55.0%

57.0%

59.0%

61.0%

95%
93%
91%
89%
87%

11.8
9.8
7.8
5.8
3.8

12.8
10.8
8.8
6.9
4.9

13.8
11.9
9.9
7.9
5.9

14.9
12.9
10.9
8.9
6.9

15.9
13.9
11.9
9.9
7.9

 

Refer to source for three other Sensitivity Analyses on Kerry's National Vote.

Required Bush share of returning Gore voters: 2000-2004 uncounted vote scenarios

Most likely scenario:


•••



Full text   http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=389&topic_id=4355303&mesg_id=4355303

 

[tiptoe]

1. Vote switching can be expected on Tuesday

Read the damn OP.

[endthewar]

2. Is this an "Oh noes!!!" type of thread, a celebration thread, or just a doomsday thread?

Can't really tell.

[tiptoe]

3. it's an analysis. n/t