2010 PA, WI and IL Senate Elections: A Comparative Demographic and True Vote Analysis

22 Jun

Richard Charnin
June 20, 2011
Updated: July 13, 2015

This analysis includes unadjusted 2008 state presidential exit poll statistics.The GOP won the 2010 WI, IL and PA senate recorded votes. But did they win the True Vote?

Obama had a 56.2% recorded share in Wisconsin. He had a whopping 63.2% in the unadjusted Wisconsin exit poll. But the popular progressive Sen. Russ Feingold lost by 5% in a traditionally progressive state. Feingold had 56% in 2004. How does one explain the 2010 loss? Was it due to unverifiable touch screens (DRE) and/or the central tabulators that miscounted the optiscan ballots?

Obama had a 61.9% recorded share in Illinois. He had a whopping 66.3% in the unadjusted exit poll. But the progressive Democrat Giannoulias lost by 2%. He led the Registered Voter (RV) polls by 42-38 and the Likely Voter (LV) polls by 43-42. He won the unadjusted exit poll by 51.0-47.0 (144,000 votes), Was the loss due to unverifiable DREs and rigged central tabulators that miscounted the optiscan ballots?

Obama had a 54.5% recorded share in Pennsylvania. He had a whopping 63.8% in the unadjusted PA exit poll. But the progressive Democrat Sestak lost by 2% He led the RV polls by 47-43 and trailed the LV polls by 49-45. He led the exit poll at 10:15pm but fell behind at 1:17am as the poll was being matched to the vote with no change in respondents. The unadjusted exit poll was a virtual 49.8-49.9% tie. Was the loss due to unverifiable DREs and rigged central tabulators that miscounted the optiscan ballots?

In 2008, Oregon voted 56.7% for Obama, nearly matching the OR True Vote Model. He had 58.2% in the exit pollster telephone poll. As a battleground state, Oregon should have been representative of the national electorate. Senator Ron Wyden led by a steady 20% in pre-election Likely Voter (LV) polls and won a 57% recorded share, matching Obama’s share.

Each of these battleground states shifted to the GOP and Wyden won Oregon in a landslide. Why did they differ from Oregon? Was it because Oregon’s recorded vote reflected the true intent of the voters? Was it due to the fact that Oregon is a 100% paper ballot state? Or that Oregon mandates hand-counts of randomly selected counties – a clear deterrent to election fraud?

Oregon was the only battleground state in which Kerry’s vote share exceeded Gore’s 2000 share. Since the 2000 election, Oregon’s recorded vote share has consistently matched the pre-election polls and the unadjusted national exit polls.

Table 1 is a comparison of the Wisconsin, Pensylvania and Illinois Final Exit Polls (i.e. recorded vote) and the True Vote Model. Note that all Final Exit Poll demographics understated the Democratic share as they were forced to match the recorded vote.

These are the key results:
– Final exit polls were forced to match the recorded vote by assuming that nearly one in six returning Obama voters defected to the GOP.
– Vote shares were n/a for the 6% that were returning 2008 third party (“Other”) and new (“DNV”) voters.
– The percentage mix of returning third-party (“Other”) 2008 voters invariably indicated that there were more returning third-party voters voters than actually voted in 2008.

In the three elections, the Democrats…
– lost the average recorded vote by 2.4%.
– won the True Vote by 3.2% (even assuming a 10% GOP edge in 2008 voter turnout).
– won the “When Decided” category by 50.0-47.6%.
– led the GOP in Party ID by 40.3-36.0%.

As always, exit poll weightings and vote shares for all categories were rounded to the nearest 1%. There is no reason why they are not displayed to the nearest tenth of a percent – unless the pollsters want to fudge the vote shares to match the recorded vote – which they do anyway. After all, 17,836 voters were interviewed in the National Exit Poll and approximately 2000 in each state. Analysts want to see the unadjusted, “pristine” numbers – but the NEP won’t allow that.

The conventional wisdom is that the 2010 midterms were a GOP blowout of epic proportions – even bigger than 1994. The party in power nearly always loses seats in the midterms. The unconventional wisdom is that the Democrats do significantly better than the recorded vote indicates in every election. There is no reason to suspect that 2010 was any different.

The media pundits accept the recorded vote and final exit polls as gospel; they never question the official results. And they never mention the fact that it is standard operating procedure for the exit pollsters to force all final national and state exit polls to match the recorded vote. This is the key “tell”. It makes no sense because the assumption is always ZERO fraud. Now how ridiculous is that?

Historically, the strong correlations between a) pre-election registered voter polls and unadjusted exit polls and b) pre-election likely voter polls and final exit polls (i.e. the recorded vote), is a clear indication of election fraud. Likely voter (LV) polls are a sub-sample of registered voter (RV) polls. We have unverifiable elections and a strange reluctance of the Democratic leadership to do anything about it.

Let’s consider the 2010 senate pre-election polls. Based on 37 LV polls (the GOP led the average by 48.1-43.5%), the pre-election model predicted a 50-48 Democratic Senate.

CNN/Time provided RV and LV polling data for 18 Senate races (Table 1). The Democrats led a combination of 18 RV and 19 LV polls by 45.2-44.6% giving them a 53-45 seat majority.
The Democratic RV margin was approximately 5% higher than the LV margin.

RV polls were not listed in final polling averages. The Democrats led the average RV poll by 49.2-40.6%. They also led the corresponding LVs by 46.6-45.8%. The Democratic margin was 8% higher in the RV polls.

The 2010 Final National Exit Poll indicated that 45% of the electorate were returning Obama voters and 45% were McCain voters. Obama’s recorded vote margin was 52.9-45.6%. Of course, the pundits will claim that the 7.3% discrepancy was due to millions of unenthusiastic Democrats who did not return to vote in 2010.

The pollsters make adjustments to the number of returning voters (the “mix”) and the vote shares in order to match the vote count. Obama won the recorded vote by 9.5 million, but his True Vote margin was at least twice that. His recorded share understated his True vote share by 4-5%. If the 2010 NEP returning voter mix is adjusted to match the 2008 recorded share (53-45%), the average Democratic share is within 1% of the GOP share – and matches the pre-election RV polls. The adjusted 53% Democratic share of the 2010 electorate is 5% lower than Obama’s True share.

Table 2 is a comparison of 18 pre-election Registered (RV) and Likely Voter (LV) polls and the recorded vote. Democrats led the RV poll average by a solid 46.4-41.1%. They led the LV sub-sample by 46.6-45.8%, within 0.4% of the recorded 48.3-47.9% share.

Table 3 is a comparison between the final pre-election LV polls, the preliminary exit polls and the recorded shares. There was a 2.9% discrepancy in margin between the average recorded vote and the exit poll. The final RCP LV projected average margin exceeded the recorded margin by 2.2%.

Table 4 displays Gender vote shares in the Final exit polls.

Table 5 displays Final state exit polls with the returning voter mix adjusted to match the 2008 recorded vote. The resulting vote shares closely match the pre-election RV polls – and the True Vote.


One response to “2010 PA, WI and IL Senate Elections: A Comparative Demographic and True Vote Analysis

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: