Exit Poll Deviations vs. Early Voting

[Bill Bored]

So far, we have seen that on average there were larger deviations between the exit polls and the vote tallies in:

1. States with Republican Governors
2. States with Straight Party Voting
3. States with both.

See this thread:
<http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x347704>

We know that Straight Party Voting can be exploited to favor one candidate over another, especially in e-votin' systems that allow certain races (e.g., President) and certain candidates (e.g., Kerry) to be excluded from the straight party option. It is also possible to go one step further and include the wrong candidate (e.g., Bush) as the straight party choice for a given party (e.g., Democratic). And one more thing: depending on the machine configuration, on Diebold Touch Screens, the precise sequence the voter uses to vote for President on a straight party ticket may result in the Presidential vote being switched or not counted at all. If the voter doesn't notice any of this, he/she will cast an unintended vote or an undervote for President in the 17 states that allow Straight Party Voting.

Another way to gain an advantage is by switching candidates' names around on the ballot, without redirecting the votes in the tabulators' database. The voter sees Kerry's name on the ballot in black and white, votes for Kerry, but the vote ends up in Bush's column in the database. The reverse happens to a Bush voter. Simple. But how would a perpetrator of election fraud know in advance whether or not to switch the names on the E-Ballot on Election Day? I.e., how would he or she know that without switching the votes, Bush would lose? Well, there are demographics, polls, etc. but it could be argued that the BEST poll (and some would say the ONLY one that counts) is the Election. This is where Early Voting comes in.

Like Straight Party Voting, Early Voting makes it easier for some people to vote, but Early Voting gives the fraudster a way to know in advance how the vote will turn out on Election Day in the very same county. If the early results are viewed before election day, this could enable a fraudster to make last minute changes to the E-Ballot that will appear on the Touch Screens on Election Day and reverse the outcome of the election.

If this happened on Nov. 2, 2004, we would therefore expect to see the following pattern in states with Early Voting:

1. Kerry wins early voting while Bush wins on Election Day in the same counties.
2. Larger Exit Poll deviations in states that allowed Early Voting, particularly those with Touch Screen/DREs such as Diebold's Accuvote TS.

So if anyone is motivated please post a list of states that allowed Early Voting in 2004 and I, TIA, or another interested party will compare the average Exit Poll deviations in these states to states that had no Early Voting.

When I looked at states with Republican Govs. and those with Straight Party Voting, I had no idea in advance whether their exit poll deviations would be larger, smaller or the same in these states compared to others. It just turned out that they were in fact larger in these categories.

Anyone want to guess whether this trend will also apply in states with Early Voting? Let's find out.

[FogerRox]

1. I was in Broward County--the Kerry headquarters in Tallahase

TOld us to push early voting. Everyone should vote early.

[Old and In the Way]

2. Thanks!

You, TIA, and others are doing important work...it's appreciated.

[Kip Humphrey]

3. Texas had both and a known straight Dem ticket exploit

that denied Kerry untold votes. The straight ticket exploit involved Opti-Scans and Hart Intercivic DREs.

[Bill Bored]

4. Well, I'd like to see the rest of 'em

because oddly enough, the TX vote tally shifted in favor of Kerry. This is why I sometimes have trouble believing these exit polls. Here is a state with many of the known exploits, Bush did better there than in 2000, yet the exit polls say it was fine (within the MOE according to TIA) and the exit polls underestimated Kerry there.

Have to look at the other states, especially the swingers!

FL

[Bill Bored]

5. Probably old news but...

PCWorld.com - States Report Early E-Voting Glitches
States Report Early E-Voting Glitches. Touch-screen bugs, laptop crashes, and server problems crop up in pre-Election Day voting

<http://www.pcworld.com/news/article/0,aid,118351,00.asp>

[Bill Bored]

6. And here's the list of states:

<http://www.ncsl.org/programs/legman/elect/absentearly.htm>

The following states permit in-person early voting at election offices and, in some states, other satellite locations.

Alaska
Nebraska
Arizona
Nevada
Arkansas
New Mexico
California
North Carolina
Colorado
North Dakota*
Florida
Oklahoma
Georgia
Tennessee
Hawaii
Texas
Indiana
Utah
Iowa
Vermont
Kansas
West Virginia
Maine

* Early voting is not offered in all counties. In North Dakota, it is offered at the county election officer’s discretion.

Average Red Shift (n=23) = -1.28%.
For Dem Govs. (10) = -0.81%
For Repub Govs. (13) = -1.65%

Not bad. The Repubs had greater Red Shifts, but we know that already. I guess there's nothing to see here folks, move along now!