2010 WI Senate True Vote Model
 Share
The version of the browser you are using is no longer supported. Please upgrade to a supported browser.Dismiss

View only
 
ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
2
This is a True Vote analysis of the 2010 Wisconsin senate race. The Final 2010 Wisconsin Senate Exit Poll was forced to match the recorded vote (Johnson defeated Feingold by 52-47%). Forcing a match to the recorded vote is standard operating procedure on the part of the exit pollsters. In order to force a match in 2004 and 2008, the exit pollsters had to assume an impossible number of returning Bush voters from the previous election. I have included a comparable Oregon analysis. Oregon Senator Wyden, a popular progressive Democrat like Feingold, was re-elected with a 57% share, closely matching Obama’s 56.8% OR share.

The returning voter mix should reflect the previous election True Vote, not the recorded vote, In 2010, the Final WI senate exit poll was again forced to match the recorded vote. In the Final, 49% of the 2010 recorded vote were cast by returning Obama voters and 43% by returning McCain voters. The ratio is consistent with Obama’s 7.5% national recorded vote margin. But a WI True Vote analysis indicates that Obama won by 57-41%.

5/6/12 Update:
The unadjusted 2008 state presidential exit polls were not available when this analysis was first posted.
Obama won the unadjusted Wisconsin poll by 63-36% (n=2,545 respondents) - a 63.9% (2-party) share. He had 56.2% recorded.
The exit poll 2.43% margin of error includes a 30% "cluster effect".
Based on the MoE, there is a 97.5% probability that Obama's True Wisconsin vote share exceeded 60.77% = 63.20% - 2.43%.

Using the unadjusted polls as a basis, Feingold won by 53.2-45.6%, a 165,000 vote margin.

But even that is conservative: The calculation uses the final exit poll vote shares that were required to match the recorded vote
The shares were adjusted to favor Johnson (Feingold had just 84% of returning Obama voters, an implausibly low share).
If Feingold had a plausible 89% of Obama voters (all else equal), he would have won by 289,000 votes with a 56% total share.

Furthermore, the model assumes 76% McCain voter turnout in 2010, compared to just 68% of Obama voters.
If just 63% of Obama voters and 81% of McCain voters turned out, then all constant, Feingold wins by 52,000 votes with a 50.6% total share.

It is logical to assume that there were more returning Obama voters than indicated in the 2010 WI exit poll. Therefore, the returning Obamavoters share of the 2010 electorate was increased. Election fraud may very well have cost Feingold the election. The True Vote analysis indicates that Feingold had an approximate 50.6% vote share and 50,000 vote margin.

Vote shares were not provided for returning third party (Other) voters and new (DNV) voters in the WI Exit Poll. These categories represented 3% and 5% of the total 2010 recorded vote, respectively. In order to match the recorded vote, Johnson must have won these voters by approximately 61-35% and 64-36%, respectively. This is a RED FLAG! Obama won third party voters by 66-20% over McCain. It strains credulity that returning third party voters would switch from Obama to Johnson in those amounts.

Note the SENSITIVITY ANALYSIS which displays vote shares for various returning voter mix assumptions (scenarios) and Feingold’s share of returning McCain voters. Although the Final Exit Poll is always forced to match the recorded vote, the Base Case True Vote scenario assumes the Final Exit Poll vote shares; only the returning voter mix was adjusted. However, it is likely that the vote shares were also adjusted to force the match.

The True Vote Base Case analysis assumes a 60% turnout of living Obama voters and a 70% turnout of McCain voters and a 1.25% annual voter mortality rate. The percentage mix of returning 2008 third-party (other) voters could not have been the 3% indicated in the WI exit poll. That would mean there were 65,000 third-party voters and there were just 44,000. Therefore, the model assigned the 1.5% excess of Other voters to New/DNV (first-time voters and others who did not vote in 2008).

Feingold was a clear winner assuming that the percentage turnout of returning Obama voters was equal or greater than that of McCain voters. But it is important to keep in mind that the WI exit poll gave Feingold just 84% of returning Obama voters. This is extremely implausible. It is difficult to accept the premise that nearly one of six Obama voters defected to Johnson.

A comparison of the exit poll demographic changes from 2004 to 2010 yields some interesting results:

When Decided: Feingold was the winner in this category.
1) Feingold led the 83% of voters who decided a week prior to the election by 51-48%, and won by 50.7-48.5%.
2) He led the 68% of voters who decided a month prior by 51-48%, and won by 50.4-49.0%.

Gender: In 2004, females comprised 53% total voters; in 2010, they were just 50% of the total.
Age: Voters over 45 comprised 50% of the electorate in 2004 and 62% in 2010.
Party ID: In 2004, Democrats were 35% and GOP 38% of the electorate. In 2010, Democrats were 37%, the GOP had 36%.
Region: Milwaukee County had 16% of voters in both 2004 and 2010. Feingold had 68% in 2004 but just 61% in 2010
Southeast (incl Waukesha) had 19% of voters in 2004 and 27% in 2010. Feingold had 39% in 2004 and 34% in 2010.

Assuming that Feingold had 88% of returning Obama voters (65% turnout) and Johnson had 93% of returning McCain voters (75% turnout),Feingold was still a 50.0-48.8% winner with a 26,000 vote margin.

3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Loading...