The Final 2004 National Exit Poll switched 6.7% of Kerry responders to Bush

The Final 2004 National Exit Poll switched 6.7% of Kerry responders to Bush

Richard Charnin
Feb.21, 2012
Updated: July 23, 2015

The myth that the early 2004 exit polls were biased for Kerry is refuted by the National Exit Poll (NEP) timeline. Kerry had 51% at 4pm (8,349 respondents). His exit poll share remained constant up to the final 13,660 respondents (51.7%). The pollsters had to switch 471 (6.7%) of Kerry’s 7,064 responders to Bush in order to have the Final NEP match the recorded vote. Assuming that Kerry had 51.7% of 125.7 million votes cast, he won by nearly 6 million votes. The True Vote Model indicates that he had 53.6% and won by 10 million.

All National Exit Poll demographic crosstabs had to be adjusted to match the vote. In the  “Voted in 2000” crosstab, 43% (52.6 million) of 2004 voters were returning Bush 2000 voters. But Bush had just 50.5 million recorded votes in 2000. Approximately 2 million died prior to the 2004 election and one million did not return to vote. Therefore, there were at least 5 million PHANTOM BUSH VOTERS. The NEP also claims that just 37% of 2004 voters were returning Gore voters. That is ridiculous; Gore had 540,000 more recorded votes than Bush. Once again, the 6% red-shift rears its ugly head (43-37=6%). THIS SIMPLE LOGIC PROVES MASSIVE 2004 ELECTION FRAUD.

The 2004 preliminary and final reported national exit poll timeline was debated  in the Game thread on Democratic Underground in 2005. The final poll indicated that Bush led the full set of 13660 respondents by his recorded vote margin. But this was before the unadjusted exit polls became available which showed that Kerry led throughout the timeline. This post discusses matching the exit poll to the recorded vote.

US Count Votes did a comprehensive analysis of the 2004 exit poll discrepancies which disproved the exit pollster’s reluctant Bush responder hypothesis.

To view the 1988-2008 Unadjusted State and National Exit Poll Database click this link: https://docs.google.com/spreadsheet/ccc?key=0AjAk1JUWDMyRdFIzSTJtMTJZekNBWUdtbWp3bHlpWGc#gid=15

The source data is provided by the Roper Center UConn

The data for each election is viewed by clicking the indicated tab at the top of the screen. State exit polls are displayed in the same row as the recorded vote. The national aggregate exit poll is calculated by weighting the state exit poll shares by votes cast.

Aggregate state share = SumProduct (exit poll  (i) * pop. weight (i) ) / Total Votes Cast, i = 1,51


11/2/04 3:59pm, 8349 respondents
http://www.richardcharnin.com/US2004G_3737_PRES04_NONE_H_Data-1.pdf
Voted Mix Kerry Bush Other
DNV.. 15% 62% 37% 1%
Gore. 39% 91% 8% 1%
Bush. 42% 9% 90% 1%
Other 4% 61% 12% 27%
Total 100% 51% 47% 2%

11/2/04 7:33pm, 11027 respondents 
http://www.richardcharnin.com/US2004G_3798_PRES04_NONE_H_Data.pdf

Vote04 Mix Kerry Bush Other
DNV.. 17% 59% 39% 2%
Gore. 38% 91% 8% 1%
Bush. 41% 9% 90% 1%
Other 4% 65% 13% 22%
Total 100% 50.9% 47.1% 2.0%

http://media.washingtonpost.com/wp-srv/politics/elections/2004/graphics/exitpolls_us_110204.gif
11/3/04 12:22am, 13047 respondents
Vote04 Mix Kerry Bush Other
DNV.. 17% 57% 41% 2%
Gore. 39% 91% 8% 1%
Bush. 41% 10% 90% 0%
Other 3% 64% 17% 19%
Total 100% 51.2% 47.5% 1.3%

Unadjusted National Exit Poll, 13660 respondents (true sample)
Data Source: Roper Center UConn
Total Kerry Bush Other
13660 7064 6414 182
Share 51.7% 47.0% 1.3%

Vote04  Mix Kerry Bush Other
DNV.. 18.4% 57% 42% 1%
Gore. 38.4% 91% 8% 1%
Bush. 39.5% 10% 90% 0%
Other 3.75% 64% 17% 19%
Total 100% 51.7% 47.0% 1.3%

11/3/04 1:24pm, Final National Exit Poll, 13660 respondents (adjusted)
http://www.richardcharnin.com/US2004G_3970_PRES04_NONE_H_Data.pdf

The Final was forced to match recorded vote by switching approximately 471 (6.7%) of Kerry’s 7,064 respondents to Bush. The average within precinct exit poll discrepancy was a nearly identical 6.5%.
Final Kerry Bush Other
13660 6593 6930 137
Share 48.3% 50.7% 1.0%

Voted Mix Kerry Bush Other
DNV.. 17% 54% 45% 1%
Gore. 37% 90% 10% 0% << 37%? Gore won 2000 recorded vote by 540k!
Bush. 43% 9% 91% 0% << 43% (52.6mm)? Bush had 50.5mm recorded!
Other 3% 71% 21% 8%
Total 100% 48.5% 51.1% 0.4%

Unadjusted NEP: Gender Demographic
Gender Mix Kerry Bush Other
Male.. 46.0% 48.0% 51.0% 1.0%
Female 54.0% 55.0% 44.0% 1.0%
Total.. 100% 51.8% 47.2% 1.0%

Final Adjusted Gender (forced to match recorded vote)
Gender Mix Kerry Bush Other
Male.. 46.0% 44.0% 55.0% 1.0%
Female 54.0% 51.0% 48.0% 1.0%
Total.. 100% 47.8% 51.2% 1.0%

True Vote Model
Voted04 Mix Kerry Bush Other
DNV.. 17.0% 57.0% 41.0% 2.0%
Gore. 41.5% 91.0% 8.0% 1.0%
Bush. 38.0% 10.0% 90.0% 0.0%
Other 3.50% 64.0% 17.0% 19.0%
Total 100 53.5% 45.1% 1.4%

2008
Obama had 61% in the unadjusted National Exit Poll (17836 respondents), but just a 52.87% recorded share. The pollsters had to effectively reduce Obama's respondents from 10873 to 9430 (13.3%) in order to force the final NEP to match the recorded vote. The True Vote Model indicates that he had 58%.

Unadjusted 2008 National Exit Poll (17836 respondents)

Final NEP (Unadjusted)
Sample Obama McCain Other
17836 10873 6641 322
Share 61.0% 37.2% 1.8%

Final NEP (forced to match the recorded vote)
Sample Obama McCain Other
17,836 9,430 8,137 269
Share 52.9% 45.6% 1.5%

Unadjusted National Exit Poll (True Vote)
Voted04 Share Votes Mix Obama McCain Other
Kerry 50.2% 57.1 43.4% 89.0% 9.0% 2.0%
Bush. 44.6% 50.8 38.6% 17.0% 82.0% 1.0%
Other 5.2% 5.9 4.5% 72.0% 26.0% 2.0%
DNV.. .... 17.7 13.4% 71.0% 27.0% 2.0%
Total 100% 131.5 100% 58.0% 40.4% 1.6%
Votes .... .... 131.5 76.3 53.0 2.2%

Final National Exit Poll (forced to match recorded)
Voted04 Share Votes Mix Obama McCain Other
Kerry 42.5% 48.6 37.0% 89.0% 9.0% 2.0% << 37.0% KERRY?
Bush. 52.9% 60.5 46.0% 17.0% 82.0% 1.0% << 46.0% BUSH? 9% DIFFERENTIAL?
Other 4.6% 5.3 4.0% 72.0% 26.0% 2.0%
DNV.. .... 17.1 13.0% 71.0% 27.0% 2.0%
Total 100% 131.5 100% 52.9% 45.6% 1.5%
Votes .... .... 131.5 69.5 60.0 2.0

http://www.cnn.com/ELECTION/2008/results/polls/#USP00p1
Final National Exit Poll (forced to match recorded)
Gender Mix Obama McCain Other
Male.. 47.0% 49.0% 48.0% 3.0%
Female 53.0% 56.0% 43.0% 1.0%
Total. 100% 52.7% 45.4% 1.9%


Election Model Forecast; Post-election True Vote Model

2004 (2-party vote shares)
Model: Kerry 51.8%, 337 EV (snapshot)
State exit poll aggregate: 51.7%, 337 EV
Recorded Vote: 48.3%, 255 EV
True Vote Model: 53.6%, 364 EV

2008
Model: Obama 53.1%, 365.3 EV (simulation mean);
Recorded: 52.9%, 365 EV
State exit poll weighted aggregate: 58.0%, 420 EV
Unadjusted National Exit Poll (17836 respondents):Obama 61%-McCain 37%
True Vote Model: 58.0%, 420 EV

2012 (2-party state exit poll aggregate shares)
Model: Obama 51.6%, 332 EV (Snapshot)
Recorded : 51.6%, 332 EV
True Vote 55.2%, 380 EV

 

Obama May Win the True Vote and Still Lose: The 2012 Presidential True Vote Projection Model

The 2012 Presidential True Vote Projection Model

Updated: April 29, 2012

The systemic Election Fraud that Democrats won’t talk about could cause Obama to lose – just like Gore and Kerry. The unadjusted 2008 State exit polls indicate that Obama had 420 EV and a 58% share – exactly matching the State and National True Vote Model (TVM).

Officially, Obama had 365 EV and a 52.9% recorded share. He needs at least a 55% True Vote share to break even – if you believe the TVM and the unadjusted exit polls. The model shows how Obama could win the True Vote and still lose the election.

The model will be updated periodically to include state and national pre-election polls. A Monte Carlo electoral vote simulation has been added to calculate the probability of winning under various scenarios.

https://docs.google.com/spreadsheet/ccc?key=0AjAk1JUWDMyRdDQzLWJTdlppakNRNDlMakhhMGdGa0E#gid=14

Sensitivity Analysis

No model is complete without considering the effects of alternative input scenario assumptions. Sensitivity tables display Obama vote shares, margins and popular vote win probabilities over a range of assumptions.

The theoretical expected EV is the product sum of the state win probabilities and corresponding electoral votes.

Pollsters and pundits never consider election fraud as the cause of exit poll discrepancies. Likely Voter (LV) pre-election polls have usually been accurate predictors of the recorded vote. The polls are subsets of Registered Voter (RV) polls. The majority of potential voters excluded in the Likely Voter Cutoff Model are newly registered young Democrats.

Final Exit polls are always forced to match the recorded vote, even when the adjustments are mathematically impossible. It’s standard operating procedure.

Pre-election and exit poll samples consider prior election recorded vote trends. But the recorded votes always understate the true Democratic share. From 1988-2008, the Democrats won the average of the state presidential unadjusted exit polls by 52-42%. The recorded margin was just 48-46%.

Of the 274 state exit polls from 1988-2008, 226 (82%) shifted from the exit poll in favor of the GOP – a zero (3.7E-31) probability. Of the 274, 126 (46%) exceeded the margin of error – a zero (8E-75) probability. At the 95% confidence level, about 14 would be expected to exceed. But that’s not all. Of the 126 which exceeded the margin or error, 123 (98%) moved from the Democrat in the poll to the Republican in the vote. The probability is zero (5.4E-106.)

Election Model Forecast; Post-election True Vote Model

2004 (2-party vote shares)
Model: Kerry 51.8%, 337 EV (snapshot)
State exit poll aggregate: 51.7%, 337 EV
Recorded Vote: 48.3%, 255 EV
True Vote Model: 53.6%, 364 EV

2008
Model: Obama 53.1%, 365.3 EV (simulation mean);
Recorded: 52.9%, 365 EV
State exit poll aggregate: 58.0%, 420 EV
True Vote Model: 58.0%, 420 EV

2012 (2-party state exit poll aggregate shares)
Model: Obama 51.6%, 332 EV (Snapshot)
Recorded : 51.6%, 332 EV
True Vote 55.2%, 380 EV

 

Using True Vote Model Sensitivity Analysis to Prove that Kerry won the 2004 Election

Using True Vote Model Sensitivity Analysis to Prove that Kerry won the 2004 Election

Richard Charnin
Feb. 8, 2012
Updated: Jan. 17, 2016

It never ends. The media still wants us to believe that Bush won the 2004 election by a 3 million vote margin (50.7-48.3%). They call researchers who proved that the election was stolen “conspiracy nuts”. But they have never debunked the overwhelming evidence that the election was a massive fraud.  Even after inflating Bush exit poll vote shares and turnout, a True Vote sensitivity analysis shows that Kerry won all plausible scenarios. It’s time for the media to tell the truth. Kerry won a landslide. The election was stolen, just as like it was stolen from Gore in 2000.

The pundits resorted to claims that the exit polls “behaved badly”, Bush voters were “reluctant to be interviewed” by the exit pollster, returning “Gore voters lied” about their past vote and there was no correlation between “vote swing and the exit poll red-shift”.   US Count Votes did a comprehensive analysis of the 2004 exit poll discrepancies which disproved the exit pollster’s reluctant Bush responder (rBr) hypothesis.

A comprehensive spreadsheet analysis of  discrepancies between state and national unadjusted exit polls and recorded votes is overwhelming proof that election fraud is systemic.  The Democrats won the 1988-2008 presidential exit polls  by 52-42%, but just 48-46% in the official recorded vote.

The 2004 National Exit Poll (NEP) displayed on corporate media election websites indicates that Bush was the winner – until one takes a closer look. As we all should know by now, exit polls are always forced to match the recorded vote – come hell or high water.  In other words, they always assume zero election fraud.

The 2004 NEP indicates that 52.6 million (43%) of the 2004 electorate were returning Bush 2000 voters and 45.1 million (37%) were Gore voters. As we have shown numerous times before, this is an impossible scenario.

Bush had just 50.5 million recorded votes in 2000. Gore had 51.0 million. Approximately 5% (2.5 million) of Bush 2000 voters died, so at most 48 million returned to vote in 2004.

But 100% turnout is impossible; therefore had to be fewer than 48 million returning Bush voters. Assuming 98% turnout, 47 million returned in 2004. That is 5.6 million less than the 52.6 million indicated in the Final 2004 National Exit Poll. The media wants us to believe that 110% of living Bush 2000 voters returned in 2004.

Where did these mysterious phantom Bush voters come from? What does that tell us about the Final? And since the Final was forced to match the recorded vote, what does that tell us about the recorded vote?

The True Vote Model

Unlike the impossible Final 2004 NEP, the 2004 True Vote Model determines a feasible (i.e. mathematically possible) and plausible (likely) number of returning Bush and Gore voters. An estimated 98% of living 2000 voter turned out in 2004.

Note: There were 6 million uncounted votes in 2000 (approximately 75% for Gore). Therefore, Gore’s True Vote margin was at least 4 million. But we will be conservative in assuming that he won by just 540,000 recorded votes.

Even if we use the bogus 2000 recorded vote  as a basis for returning Bush and Gore voters and apply 12:22am NEP vote shares, Kerry is the clear winner of the Base Case scenario. He has 52.2% and a 7.3 million vote margin – with a 97% win probability.

Note that Kerry won the unadjusted National Exit Poll (13660 respondents) with a 51.7% share. The Final NEP (also 13660 respondents) has Bush winning 50.7-48.3% (the recorded vote).

Let’s use the 2000 unadjusted state exit poll aggregate (Gore had 50.8% and won by 6 million) as a basis for calculating the 2004 True Vote. Kerry’s True Vote is 54.3%. He wins the base case scenario by 12.8 million votes with a 100% win probability.

The base case assumes an equal 98% turnout of living Bush and Gore voters. Let’s assume that only 90% of Gore voters and 98% of Bush voters return. Kerry is still the winner by 10 million with a 53.2% share. He also wins the worst case scenario by 3.8 million with 50.8%.

Exit poll naysayers insist that Kerry’s  early vote shares were inflated and that the Final shares listed on CNN should be used: Kerry’s 57% share of new voters is reduced to 54% and his 10% share of returning Bush voters cut to 9%. We’ll do better than that. 

View the True Vote Model sensitivity analysis tables. In the worst case scenario, Kerry has just 53% of new voters and 8% of Bush voters.  Kerry is still the winner by 5.9 million votes with a 51.6% share and a 94% win probability.

This refutes the media myth that Bush won. Let us count the ways:
1. Adjustments made to the National Exit Poll in order to match the recorded vote were impossible (required that 110% of living Bush 2000 voters return in 2004).
2. Kerry is a 52.2% winner assuming 98% of living Bush and Gore 2000 voters turned out in proportion to the 2000 recorded vote.
3. Kerry is a 54.3% by 12.8 million votes assuming the 2000 True Vote as the basis for returning voters.
4. Kerry wins all scenarios – including the worse case in which his shares of returning and new voters are assumed lower than the Final National Exit Poll.
5. Even assuming 98% Bush / 90% Gore turnout, Kerry wins all scenarios.

A statistical analysis of 49 Ohio 2004 exit poll precincts was produced by Ron Baiman and Kathy Dopp at U.S. Count Votes.

Over 40% of Ohio’s exit polled precincts had statistically significant discrepancies. This is over four times the number of expected precincts with significant discrepancy.
• 45.1% (22 of 49) of Ohio’s polled precincts have significant discrepancy when calculations assume that official vote counts most accurately estimate actual vote share, and
• 40.7% (20 of 49) of Ohio’s polled precincts have significant discrepancy when calculated by assuming that exit poll results are a better estimate of real vote share.

Ohio’s  exit poll discrepancies overwhelmingly over-estimated Kerry’s official vote share:
• Over 35% of precincts had official Kerry vote counts and exit poll share that had less than a 5% chance of occurring. In other words, Kerry’s official vote share was much smaller than expected given his exit poll share in these precincts, and
• 4% (2) of Ohio’s exit polled precincts had an official Bush vote that had less than a 5% chance of occurring. In these precincts Bush official vote share (assumed to be one minus their Kerry share) was much smaller than expected, given Bush’s exit poll share.

RFK Jr wrote the landmark article: Was the 2004 Election Stolen

Will any media pundits, election analysts or political scientists  come  forward to refute the massive evidence of fraud and prove that Bush really did win a fair election? Don’t hold your breath. Job tenure is everything.

Simulation forecast trends are displayed in the following graphs:

State aggregate poll trend
Electoral vote and win probability
Electoral and popular vote
Undecided voter allocation impact on electoral vote and win probability
National poll trend
Monte Carlo Simulation
Monte Carlo Electoral Vote Histogram

Election Model Forecast; Post-election True Vote Model

2004 (2-party vote shares)
Model: Kerry 51.8%, 337 EV (snapshot)
State exit poll aggregate: 51.7%, 337 EV
Recorded Vote: 48.3%, 255 EV
True Vote Model: 53.6%, 364 EV

2008
Model: Obama 53.1%, 365.3 EV (simulation mean);
Recorded: 52.9%, 365 EV
State exit poll aggregate: 58.0%, 420 EV
True Vote Model: 58.0%, 420 EV

2012 (2-party state exit poll aggregate shares)
Model: Obama 51.6%, 332 EV (Snapshot)
Recorded: 51.6%, 332 EV
True Vote Model: 55.2%, 380 EV

 

Voting Early (Paper Ballots) vs. Election Day (Machines)

Voting Early (Paper Ballots) vs. Election Day (Machines)

This 2008 analysis compares exit poll discrepancies in states that voted early by mail or hand-delivered paper ballots. Approximately 30% of the 131 million total votes were cast early. The exit poll red-shift to the GOP is negatively (-0.50) correlated to early mail or in-person voting (paper ballot). In other words, the unadjusted exit polls are a closer match to the recorded vote in early-voting states where, presumably, election fraud is minimal.

In general, exit poll discrepancies from the recorded vote (red-shift) are lower in states with a high percentage of early paper ballot voting. Conversely, states that utilize unverifiable DREs on Election Day have much higher exit poll discrepancies – as one would intuitively expect.

The 15 states with the highest early voting turnout had an average 2.3% red-shift. The 15 with the lowest early turnout had an average 6.8% red-shift.

For example, the states with the highest percentage of early/hand-delivered paper ballots early/hand-delivered paper ballots had tiny red-shifts (Pct,R/S): OR (100%,1.75%), WA (89%,0.54%) and CO (79%, -1.8%).

This scatter-chart shows that as the percentage of early (Vote-by-mail or hand-delivered) paper-ballots increase, the exit poll red-shift decreases. Note that the three points at the extreme right represent CO, WA, OR.

Approximately 30% of votes cast were mailed or hand-delivered and 7% of paper ballots were recorded late (absentee, provisional, etc.). The remaining 63% that were recorded on Election Day were a combination of DREs, Optical scanners and punch card machines. Since 30% of total votes cast in 2008 were on unverifiable DREs, then about 50% of Election Day voting was on DREs. And that explains why exit poll discrepancies were highest in states that only had Election Day voting.

Now what about the votes recorded AFTER Election Day – the Late (paper ballot) votes? How did the Democratic Late Vote share compare to the overall recorded vote? Not surprisingly, since late votes were cast on paper ballots (provisional, absentee, etc.), the Democrats did much better.

Proof: there were 121 million votes recorded before or on Election Day. Obama had 52.4%. But he had 59.2% of 10 million late recorded votes.

Here is the takeaway: If you have the option, vote early using paper ballots. Don’t wait until Election Day to vote in cyberspace. And lobby election officials to mandate that, at minimum, the paper ballots are hand counted in randomly selected precincts or counties.

Election activists who are opposed to voting early by mail or hand-delivered paper ballots should check out Oregon, Washington and Colorado. Oregon installed its vote-by-mail system in 1998. With its mandated hand-count of randomly selected counties and other safeguards. Since 2000, Oregon has by far the best record of all the battleground states based on various statistical measures of accuracy. Washington and Colorado have recently followed suit. Is it just coincidental that the three states with the highest early voting rates had the lowest exit poll discrepancies?

Election Model Forecast; Post-election True Vote Model

2004 (2-party vote shares)
Model: Kerry 51.8%, 337 EV (snapshot)
State exit poll aggregate: 51.7%, 337 EV
Recorded Vote: 48.3%, 255 EV
True Vote Model: 53.6%, 364 EV

2008
Model: Obama 53.1%, 365.3 EV (simulation mean);
Recorded: 52.9%, 365 EV
State exit poll aggregate: 58.0%, 420 EV
True Vote Model: 58.0%, 420 EV

2012 (2-party state exit poll aggregate shares)
Model: Obama 51.6%, 332 EV (Snapshot)
Recorded: 51.6%, 332 EV
True Vote: 55.2%, 380 EV