WI2012MuniRecall
The version of the browser you are using is no longer supported. Please upgrade to a supported browser.Dismiss

ABCDEFGHIJKLMNOPQRST
1
2
The Wisconsin 2012 Recall Database True Vote Model
3
4
richardcharnin.com
5
6
http://richardcharnin.wordpress.com/
7
8
http://richardcharnin.wordpress.com/2012/07/24/the-walker-recall-municipal-database-a-true-vote-model/
9
10
7/28/2012 - added GAB Ward by Ward results
11
8/5/2012 - added Winnebago TSX vs, Optiscan probability analysis
12
13
Using the Model
14
15
There are three voting units: 1- county, 2-municipality, 3- Unit /ward
16
1. To run the model for a county, view the code numbers on the "Input" sheet. They were copied from the "Muni" sheet to save time.
17
2. To run the model for a municipality, scroll the "Muni" sheet to locate the Muni row number
18
3. To run the model for a Unit/ward, scroll the "WardData" sheet to locate the Unit/ward row number
19
4. Go to the Input sheet and enter the row number in cell B7 for a county or muni. For a ward/unit, enter the row number in B8.
20
21
If Input cells B7and B8 both contain row numbers, the Ward True Vote is calculated.
22
To calculate the True Vote for a county or muni, leave B8 blank.
23
24
It is important to note that although most units (ward groups) match exactly in 2008 and 2012, there are many that do not. Therefore, a direct comparable analysis is impossible. I consolidated some units to make them more comparable. Waukesha, Kenosha and La Crosse counties are particularly difficult groups that do not match from 2008 to 2012.
25
26
Check the 2012/2008 turnout percentage in column T of "WardData". If it is not between 50-120%, then it is probably not a comparable match.
27
If the unit wards do not match, analyze the municipality.
28
29
Barrett's share of Obama voters is calculated automatically using a county partisanship increment relative to Obama's recorded share.
30
For example, Barrett's share of returning Obama voters is adjusted upward from 90% to 95% in Dane and reduced to 84% in Waukesha.
31
Vote shares set in the 'Muni' screen apply equally to all munis (i.e. Barrett's 90% share of Obama voters and 7% share of McCain voters).
32
33
According to the WI 2010 Exit Poll (forced to match the recorded vote), Barrett had 83% of returning Obama voters and 7% of McCain voters.
34
Barrett's shares of returning Obama and McCain voters can be set as defaults on the Input sheet. Walker's shares are adjusted automatically.
35
36
Although the defaults can be overridden, it should not be necessary. View the SENSITIVITY ANALYSIS tables on the "Input" sheet to save time and effort. Nine scenarios based on vote shares and nine on turnout are calculated automatically for the selected county, muni, or ward unit.
37
38
Key Muni statistics:
39
Barrett True Vote - a function of voter turnout and shares of returning and new voters
40
Voter Turnout - defined as 2012 voters/living 2008 voters
41
Required Walker share of Obama voters to match the recorded vote - given 2008 voter turnout and assuming he had 95% of McCain voters,
42
Red-shift: change from True Vote to recorded vote
43
44
INPUT DATA, TRUE VOTE MODEL and SENSITIVITY ANALYSIS
45
46
Estimated Barrett share of
47
returning Obama (default 2008 recorded adjusted for partisanship)
48
returning McCain (default 7%)
49
New voters (default 2008 Obama share)
50
51
2008 Voter Turnout in 2012
52
Obama; default set to total 2012 turnout if left blank
53
McCain: default set to total 2012 turnout if left blank
54
Note:
55
2012 Total Turnout: Must be set low enough to return a positve number of new voters.
56
New voters = total voters - returning voters
57
58
Default Barrett Wisconsin shares of:
59
Obama
60
McCain
61
New
62
63
Sensitivity Analaysis Table Increment (above and below the base case):
64
Barrett shares of returning Obama and McCain voters
65
2012 Turnout of Obama and McCain voters
66
67
Voter Mortality
68
69
70
DATA AND ANALYSIS SHEETS
71
72
73
Counties/graphs
74
The 72 counties are displayed in the same format as on the Muni sheet
75
The table can be sorted by any column.
76
Input data assumptions apply to all counties.
77
78
Muni
79
Voting data by Munis with True vote shares and other stats (protected)
80
81
MuniSort
82
Voting data by Munis with vote shares. Unprotected for sorting.
83
84
WardData
85
Voting data by Wards with vote shares and other stats (protected)
86
87
WardSort
88
Recall Vote data by Wards pre-sorted by counties/vote shares (protected)
89
90
2010 Exit Poll
91
Recorded vote vs. True Vote Demographic crosstabs
92
93
2012-2010
94
Vote change analysis (unprotected for sorting).
95
96
Machines
97
Analysis of Voting methods by county
98
99
100
Wisconsin True Vote Model