Reply #202: I've been racking my brain on NC too: great job!

[chomskysright]

Hiya: great job. I've been looking at NC data too.

You're way ahead of me. I was looking at the percentage of NC registered Dems (46%, 2004) and registered Repub (34%) and the rest independent and trying to get a grip on the cross-voting mechanism in order to see that the exit poll info from the major stations was too good of a match with the data put out by the BOE.

I've just sent your e mail over to several newspaper people: these folks: Cecil Bothwell, in particular, had a court order pending in order to get absentee ballots in Buncombe:

here is what I have been trying to get a wrap on:

The CNN/ ABC/ NBC/ FOX national poll was on a quantitatively/ qualitatively different sample that the population as measured by the BOE NC:

here is what I've been working with:

COMPARISON OF POPULATION OF VOTE PER NC BOE AS OF 11.12.04 ( n = 3500998 )

AND EXIT POLL OF 2167 PARTICIPANTS ON CNN WEB PAGE ( n = 2167)

Assumptions: two comparable populations
Problem: the CNN sample was significantly (have not measured this statistically but eye-balling the matter it appears to be statistically significant) more Republican than the BOE data re: party affiliation.

(perhaps not useful): across the state there was about a 9% increase in registered voters pertaining to party affiliations: Item 3: percentage increase of registered voters by partyfrom October 23, 2003 to November 6

Democrats: .907 increase or approx 9 %

Republicans: .914 increase or approx 9%

Libertarian: .746 increase or approx 7.5%

Unaffiliated: .836 increase or approx 8 and one-third percent


Hypothesis: the actual vote data is not significantly different from the exit poll data which appears to be significantly different from the BOE voter registration affiliation which is current as of 11.6.04.

The contest data below was updated on 11/12/2004 3:31:15 PM. This is actual vote

Repub: 1,961,188 percentage of total vote taken: .560 or 56% Does this reflect a cross over (Democratic to Republican AND UA to Republican) of approx 12% as would be assumed to happen as associated with the poll exit data?

Dem 1,525,821 percentage of total vote taken: .435 or 43.5 %


There are termed here to be UA or unaffiliatedLIBERTARIAN: 11,731 and NADER: 1802 and BROWN : 348 and COBB: 108

TOTAL: 3,500,998

NOW, in state, per NC BOE, party affiliation (11.6.04 NC BOE web page): This is voter registration data

Repub: 1,903,333 34% state Republicans registered

Dem: 2,582,712 46% state Dems per voter registration affiliation

Libertarian12762 Unaffiliated 1,021,701
18.5 % state Unaffiliated & Libertarians 1.5%

Total for Dems plus Libertarians / UA = 3,617,175; 12% cross-over, as postulated by template of exit-poll which is
skewed toward Republicans-- 434,061- What it seems we don’t know is what is the percentage of the registered voters----of Dems / Repub/ Other -----voted?

Now, exit poll data, Vote by Party ID per CNN: (with gathered data apparently skews toward Republicans given BOE registration; however we do not know how the CNN affiliated pollsters asked about their party affiliation e.g., :”Are you a Republican” will get a different answer than “Did you vote Republican”; We will assume that this is a population skewed in terms of being more Republican given the BOE data: what would be the impact on the exit poll? Assumably: less cross-over to Kerry from Republicans and more cross over to Bush from Democrats. One might also assume that there were two questions asked: “How are you registered” and then “How did you vote” or they would not have been able to present this. So, if we assume that the sample was representative of the state of NC (which it does not appear to be given the BOE voter registration data)

Repub (40%) 96% BUSH 4 % KERR per exit poll, then cross voting of Republicans for Dems = 4%

Democrat (39%) 16% BUSH 84% KERRy per exit poll, then cross voting of Democrats for Bush = 16%

So, assuming that the population is ‘intact’ the overall cross—voting in the direction of the Republicans could be estimated to be 12%


Independent (21%) 56% -3 41% Problems? No 1: (This only equals 97%; what gives?: assume these votes went to UA person) So, 9% counted toward Bush if assuming that UA picked up the other 3% which is not verified)



Repub:
Item 6: comparison of registered voters, by party, as associated with 11.12.04 data:

a. percentage of registered Democrats that voted:

59%

b. percentage of registered Republicans that voted:

97%